Report

Stereoselective Bimolecular Phenoxy Radical Coupling by an Auxiliary (Dirigent) Protein Without an Active Center

See allHide authors and affiliations

Science  17 Jan 1997:
Vol. 275, Issue 5298, pp. 362-367
DOI: 10.1126/science.275.5298.362

Abstract

The regio- and stereospecificity of bimolecular phenoxy radical coupling reactions, of especial importance in lignin and lignan biosynthesis, are clearly controlled in some manner in vivo; yet in vitro coupling by oxidases, such as laccases, only produce racemic products. In other words, laccases, peroxidases, and comparable oxidases are unable to control regio- or stereospecificity by themselves and thus some other agent must exist. A 78-kilodalton protein has been isolated that, in the presence of an oxidase or one electron oxidant, effects stereoselective bimolecular phenoxy radical coupling in vitro. Itself lacking a catalytically active (oxidative) center, its mechanism of action is presumed to involve capture of E-coniferyl alcohol-derived free-radical intermediates, with consequent stereoselective coupling to give (+)-pinoresinol.

Bimolecular phenoxy radical coupling is involved in numerous biological processes, including lignin (1), lignan (2, 3), and suberin (4) biosynthesis in vascular plants, fruiting body development in fungi (5), and insect cuticle melanization and sclerotization (6), as well as in the formation of aphid pigments (7) and algal cell wall polymers (8).

In contrast to the marked specificity observed for these varied biological systems, all previously described chemical (9) and enzymatic (10) bimolecular phenoxy radical coupling reactions in vitro have lacked strict regio- and stereospecific control. That is, if chiral centers are introduced during coupling in vitro, the products are racemic, and different regiochemistries can result if more than one potential coupling site is present. Thus, the ability to generate a particular enantiomeric form or a specific coupling product in vitro is not under explicit control. Nonetheless, bimolecular phenoxy radical coupling in vivo can lead to well-defined biopolymers and oligomers, such as melanins, lignins, and lignans, although the mechanism has been unclear. The matter is further complicated because a large number of oxidative enzymes with broad substrate specificity that exist in nature have been attributed narrow physiological functions. For example, in lignification, some six distinct oxidases (1, 11), including peroxidases and laccases, have been assigned roles in lignin synthesis based on their abilities to oxidize monolignols (lignin precursors). Accordingly, one-electron oxidation of the monolignol, E-coniferyl alcohol, results in “random” bimolecular radical coupling to afford initially dimeric products, such as (±)-dehydrodiconiferyl alcohols, (±)-pinoresinols, and (±)-guaiacylglycerol 8-O-4′-coniferyl alcohol ethers (Fig. 1A). Further oxidative coupling with monolignols then gives rise to the macromolecular lignins. It is inconceivable, however, that lignin formation would be left to the vagaries of such a wide range of enzymes, or be realized in a haphazard manner.

Fig. 1.

Bimolecular phenoxy radical coupling products from E-coniferyl alcohol. (A) Dimeric lignans formed via “random” coupling. (B) Stereoselective coupling to give (+)-pinoresinol. (C and D) HPLC profiles show chirality of pinoresinol obtained for each case, respectively. [See (16) for elution details.]

In addition to lignins, vascular plants contain a widely distributed, structurally diverse class of dimeric phenylpropanoid products known as lignans (2, 12). They are considered to arise via bimolecular phenoxy radical coupling (13) but under conditions where both the regio- and stereochemistries are explicitly controlled in order to account for their observed optical activities. Significantly, only a relatively small number of different bimolecular coupling modes are observed, with the 8,8′-linkage being the most prevalent (2, 12).

To confer stereospecificity in 8,8′-linked lignan formation, we have found that a coupling agent, a 78-kD protein, is involved. This protein has no detectable catalytically active oxidative center and apparently serves only to bind and orientate the coniferyl alcohol-derived free radicals, which then undergo stereoselective coupling. The formation of free radicals, in the first instance, requires the oxidative capacity of either a nonspecific oxidase or even a nonenzymatic single-electron oxidant.

Initial assays in which crude “cell wall preparations” from Forsythia suspensa were used (3, 14) revealed that entry into the various 8,8′-lignan skeleta occurs by coupling two achiral molecules of E-coniferyl alcohol to give (+)-pinoresinol (Fig. 1B). Both radical intermediate species, presumed bound and orientated at the supposed “(+)-pinoresinol synthase” (+PS) active site, approach each other from their si faces.

Significant difficulties were encountered in the solubilization of the putative +PS but these were overcome by using a potassium phosphate buffer extraction to remove readily soluble proteins from the cell-wall enriched homogenate. The remaining residual plant debris was consecutively extracted first with chilled acetone at −20°C and then potassium phosphate buffer containing 1% Triton X-100. After such treatments, the +PS activity was readily solubilized in 1 M NaCl (15).

Precipitation of the +PS by ammonium sulfate (40 to 80% saturation) gave a preparation that was subjected to cation exchange [MonoS and perfusion (POROS SP-M)] and gel filtration (S200) chromatography (15). In the initial MonoS chromatographic step, several oxidases were first eluted, all of which catalyze nonspecific oxidations of E-[9-3H]coniferyl alcohol leading to racemic dimers. In contrast, fractions capable of engendering (+)-pinoresinol formation eluted later, when 333 mM Na2SO4 in 40 mM MES-NaOH buffer (pH 5.0) was used as eluent. These fractions were combined and applied to a POROS SP-M column, the elution from which with a linear gradient of Na2SO4 (0 to 0.7 M) gave four overlapping fractions (I to IV) as shown (Fig. 2A).

Fig. 2.

Fractionation of protein mixture catalyzing (+)-pinoresinol formation by perfusion (POROS SP-M) chromatography. (A) Separation of proteins into four overlapping fractions I-IV, (B) purified fraction I, and (C) purified fraction III. [See (15) for elution details.]

The four fractions (I to IV) from the POROS SP-M chromatographic step were individually rechromatographed (see Fig. 2, B and C, for profiles of fractions I and III, respectively), each being subsequently assayed for +PS activity with E-[9-3H]coniferyl alcohol as substrate for 1 hour (16). Fraction I had very little +PS activity (< 5% of total activity loaded onto the POROS SP-M column), whereas fraction III catalyzed nonspecific oxidative coupling to give the (±)-dehydrodiconiferyl alcohols, (±)-pinoresinols, and (±)-erythro/threo guaiacylglycerol 8-O-4′-coniferyl alcohol ethers displayed in Fig. 1A. When fractions I and III were combined, however, the original +PS synthesizing activity was fully restored, that is, bimolecular coupling was reestablished with complete stereoselectivity.

Subsequent gel filtration (S200) chromatography of fraction I gave a protein of native molecular weight ∼78 kD, whereas SDS-polyacrylamide gel electrophoresis showed a single band at ∼27 kD (15), suggesting that the native protein exists as a trimer. Isoelectric focusing of the native protein on a polyacrylamide gel (pH 3 to 10 gradient) revealed the presence of six bands. After isoelectric focusing, each of these bands was electroblotted onto a polyvinylidene fluoride (PVDF) membrane and subjected to amino-terminal sequencing, which established that all had similar sequences indicating a series of isoforms. The ultraviolet-visible spectrum of the protein had only a characteristic protein absorbance at 280 nm with a barely perceptible shoulder at ∼330 nm (15). Inductively coupled plasma (ICP) analysis gave no indication of any metal being present in the protein. Thus, the 78-kD protein lacks any detectable catalytically active (oxidative) center.

Attention was next directed to the oxidase preparation (fraction III). Although not purified to electrophoretic homogeneity, the electron paramagnetic resonance (EPR) spectrum of this protein preparation resembled that of a typical plant laccase. We then studied the fate of E-[9-3H]coniferyl alcohol (2 μmol ml−1, 14.7 kBq) in the presence of, respectively, the auxiliary oxidase (fraction III, Fig. 3A), the 78-kD protein (Fig. 3B), and both fraction III and the 78-kD protein together (Fig. 3C) (16). With the fraction III preparation alone, only nonspecific bimolecular radical coupling occurs to give the (±)-dehydrodiconiferyl alcohols, (±)-pinoresinols, and (±)-erythro/threo guaiacylglycerol 8-O-4′ coniferyl alcohol ethers depicted in Fig. 1A. With the 78-kD protein by itself, however, a small amount of (+)-pinoresinol formation (<5% over 10 hours) was observed, this being presumed to result from residual traces of oxidizing capacity in the preparation (see below).

Fig. 3.

Time courses for E-coniferyl alcohol depletion and formation of corresponding lignans during incubation in presence of: (A) fraction III (12 μg protein ml−1); (B) dirigent protein (770 pmol ml−1); and (C) fraction III (12 μg protein ml−1) and dirigent protein (770 pmol ml−1) together. ˆ, coniferyl alcohol (calculated as dimer equivalents); •, (+)-pinoresinol; ▴, (±)-pinoresinols; □, (±)-dehydrodiconiferyl alcohols; ▵, (±)-erythro/threo guaiacylglycerol 8-O-4′-coniferyl alcohol ethers; ⋄, total of all lignans. [Values in (C) are corrected for +PS activity noted in (B), that is, <5% over 10 hours.]

When both fraction III and the 78-kD protein were combined, full catalytic activity and regio- and stereospecificity in the product was reestablished, whereby essentially only (+)-pinoresinol was formed. Note also that with fraction III alone, and when fraction III was combined with the 78-kD protein, the rates of substrate depletion and dimeric product formation were nearly identical. Moreover, essentially no turnover of the dimeric lignan products occurred in either case in the presence of the oxidase, during the time period (8 hours) examined (Fig. 3, A and C): subsequent dimer oxidation does not occur when E-coniferyl alcohol, the preferred substrate, is still present in the assay mixture. The 78-kD protein therefore appears to determine the specificity of the bimolecular phenoxy radical coupling reaction. We thus propose to describe this new class of proteins as dirigent proteins (Latin: dirigere, to align or guide). Gel filtration studies were also carried out with mixtures of the dirigent and fraction III proteins, in order to establish whether any detectable protein-protein interaction might account for the stereoselectivity, but no evidence in support of complex formation (that is, to higher molecular size entities) was observed.

We then determined the effect that the dirigent protein would have on plant laccase-catalyzed monolignol coupling. E-[9-3H]Coniferyl alcohol (4 μmol ml−1, 29.3 kBq) was incubated with a 120-kD laccase (previously purified from Forsythia intermedia stem tissue) over a 24-hour period, in the presence and absence of the dirigent protein (16). As before, incubation with laccase alone gave only racemic dimeric products, with (±)-dehydrodiconiferyl alcohols predominating (Fig. 4A). In the presence of the dirigent protein, however, the process was now primarily stereoselective, affording (+)-pinoresinol (Fig. 4B), rather than being nonspecific as observed when only laccase was present. The rates of both E-coniferyl alcohol (substrate) depletion and the formation of the dimeric lignans, respectively, were similar with and without the dirigent protein (17). Notably, when the oxidizing capacity (that is, laccase concentration) was lowered fivefold, only (+)-pinoresinol formation was observed. Thus, complete stereoselectivity is preserved when the oxidative capacity does not exceed a point where the dirigent protein is saturated. Assays were also conducted with E-[9-2H2, OC2H3]coniferyl alcohol and the dirigent protein in the presence of laccase (18). After incubation, the newly formed pinoresinol was consecutively purified by reversed-phase and chiral column high-performance liquid chromatography (HPLC), with the eluent from the latter subjected to mass spectrometric analysis. Liquid chromatography-mass spectrometry (LC-MS) analysis of the resulting (+)-pinoresinol (>99% enantiomeric excess, Fig. 5B) gave a molecular ion with a mass-to-charge ratio (m/z) of 368 (Fig. 5A), thus establishing the presence of 10 2H atoms and verifying that together the laccase- and dirigent protein-catalyzed stereoselective coupling of E-[9-2H2, OC2H3]coniferyl alcohol.

Fig. 4.

Time courses for E-coniferyl alcohol depletion and formation of corresponding lignans during incubation in presence of (A) Forsythia intermedia laccase (10.7 pmol protein ml−1) and (B) F. intermedia laccase (10.7 pmol protein ml−1) and dirigent protein (770 pmol ml−1) together. [See Fig. 3 for symbol legend; (B) is corrected for residual activity shown in Fig. 3B.]

Fig. 5.

LC/MS analysis of [9,9′-2H2, OC2H3]pinoresinol obtained after incubation of E-[9-2H2, OC2H3]coniferyl alcohol with dirigent protein (770 pmol ml−1) and laccase (4.1 pmol ml−1). (A) LC/MS fragmentation pattern of decadeuterated pinoresinol with molecular ion (m/z) = 368. (B) Total ion current showing relative ratio of (+)- and (−)-forms of pinoresinol after elution from Chiralcel OD column.

Other auxiliary one-electron oxidants can also facilitate stereoselective coupling with the dirigent protein. Ammonium peroxydisulfate readily undergoes homolytic cleavage (19) and is routinely used as a one-electron oxidant in acrylamide polymerization. Ammonium peroxydisulfate was first incubated with E-[9-3H]coniferyl alcohol (4 μmol ml−1, 29.3 kBq) for 6 hours (16). Nonspecific bimolecular radical coupling was observed, to afford predominantly (±)-dehydrodiconiferyl alcohols as well as the other racemic lignans (Table 1). However, when the dirigent protein was added, the stereoselectivity of coupling was dramatically altered to give primarily (+)-pinoresinol at both concentrations of oxidant, together with small amounts of racemic lignans. This result established that even an inorganic oxidant, such as ammonium peroxydisfulfate, could promote (+)-pinoresinol synthesis in the presence of the dirigent protein, even if it was not oxidatively as selective toward the monolignol as was the fraction III oxidase or laccase.

Table 1.

Effect of dirigent protein on product distribution from E-coniferyl alcohol oxidized by ammonium peroxydisulfate (6-hour assay).

View this table:

Next, the effects of incubating E-coniferyl alcohol (4 μmol ml−1, 29.3 kBq) with flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD) were investigated because, in addition to their roles as enzyme cofactors, they can also oxidize various organic substrates (20). Thus, E-[9-3H]coniferyl alcohol was incubated with FMN and FAD, respectively, for 48 hours (16, 21). In every instance, E-coniferyl alcohol oxidation was more rapid in the presence of FMN (Fig. 6A) than FAD (Fig. 6C). Although these differences between the FMN and FAD catalyzed rates of E-coniferyl alcohol oxidation were not anticipated, a consistent pattern was sustained: racemic lignan products were obtained, with the (±)-dehydrodiconiferyl alcohols predominating as before. When the time courses were repeated in the presence of the dirigent protein, a dramatic change in stereoselectivity was observed (Fig. 6, B and D), where essentially only (+)-pinoresinol formation occurred. Again, the rates of E-coniferyl alcohol depletion, when adjusted for the traces of residual oxidizing capacity (<5% over 10 hours) in the dirigent protein preparation, were dependent only upon [FMN] and [FAD], as were the total amounts of dimers formed. When full depletion of E-coniferyl alcohol occurs, the corresponding lignan dimers can begin to undergo oxidative changes as a function of time; specifically, FMN is able subsequently to oxidize pinoresinol, in open solution, after the E-coniferyl alcohol has been fully depleted.

Fig. 6.

Time courses for E-coniferyl alcohol depletion and formation of corresponding lignans during incubation in presence of (A) FMN (0.5 μmol ml−1), (B) FMN (0.5 μmol ml−1) and dirigent protein (770 pmol ml−1) together, (C) FAD (0.5 μmol ml−1), and (D) FAD (0.5 μmol ml−1) and dirigent protein (770 pmol ml−1) together. [See Fig. 3 for symbol legend; (B) and (D) are corrected for residual activity shown in Fig. 3B.]

We found that the coupling stereoselectivity was substrate specific. Neither E-p-[9-3H]coumaryl (4 μmol ml−1, 44.5 kBq) or E-[8-14C]sinapyl alcohols (4 μmol ml−1, 8.3 kBq),

which differ from E-coniferyl alcohol only by a methoxyl group substituent on the aromatic ring, yielded stereoselective products when incubated for 6 hours with FMN and ammonium peroxydisulfate, respectively, in the presence and absence of the dirigent protein (22). E-Sinapyl alcohol readily underwent coupling to afford syringaresinol, but chiral HPLC analysis revealed that the resulting products were, in every instance, racemic (Table 2). Interestingly, by itself the 78-kD dirigent protein preparation catalyzed a low level of dimer formation, as previously noted, but only gave rise to racemic (±)-syringaresinol formation, which is presumably a consequence of the residual traces of contaminating oxidizing capacity present in the protein preparation. In an analogous manner, no stereoselective coupling was observed with E-p-coumaryl alcohol as substrate. That is, only E-coniferyl alcohol undergoes stereoselective coupling in the presence of the dirigent protein. The low level of racemic syringaresinols obtained with the dirigent protein preparation alone confirms that traces of contaminating oxidase activity were present. Given the marked substrate specificity of the dirigent protein for E-coniferyl alcohol, it will be of considerable interest to determine how it differs from that affording (+)-syringaresinol in Eucommia ulmoides (23).

Table 2.

Effect of dirigent protein on coupling of E-sinapyl alcohol (6-hour assay).

View this table:

In regard to a mechanism for stereoselective coupling, three distinct possibilities can be envisaged. The most likely is that the oxidase or oxidant generates free-radical species from E-coniferyl alcohol, and that the latter are the true substrates that bind to the dirigent protein prior to coupling. The other two possibilities would require that E-coniferyl alcohol molecules are bound and orientated on the dirigent protein, thereby ensuring that only (+)-pinoresinol formation occurs upon subsequent oxidative coupling: this could occur either if both substrate phenolic hydroxyl groups were exposed so that they could readily be oxidized by an oxidase or oxidant, or if an electron transfer mechanism were operative between the oxidase or oxidant and an electron acceptor site or sites on the dirigent protein.

Among the three alternative mechanisms, three lines of evidence suggest “capture” of phenoxy radical intermediates by the dirigent protein. (i) The rates of both substrate depletion and product formation are largely unaffected by the presence of the dirigent protein. If capture of the free-radical intermediates is the operative mechanism, then the dirigent protein would only affect the specificity of coupling when single-electron oxidation of coniferyl alcohol is rate determining. (ii) An electron transfer mechanism is currently ruled out, because we observed no new ultraviolet-visible chromophores in either the presence or absence of an auxiliary oxidase or oxidant, under oxidizing conditions. (iii) Preliminary kinetic data (Table 3) support the concept of free-radical capture based on the formal values of Michaelis constant (Km) and maximum velocity (Vmax) characterizing the conversion of E-coniferyl alcohol into (+)-pinoresinol, with the dirigent protein alone and in the presence of the various oxidases or oxidants (24). If free-radical capture by the dirigent protein is the operative mechanism, the Michaelis-Menten parameters obtained will only represent formal rather than true values, because the highest free-energy intermediate state during the conversion of E-coniferyl alcohol into (+)-pinoresinol is still unknown and the relation between the concentration of substrate and that of the corresponding intermediate free radical in open solution has not been delineated.

Table 3.

Effect of various oxidants on formal Km and Vmax values for the dirigent protein (770 pmol ml−1) during (+)-pinoresinol formation from E-coniferyl alcohol.

View this table:

Bearing these qualifications in mind, we estimated formal Km and Vmax values for the dirigent protein preparation. As noted earlier, it was capable of engendering formation of low levels of both (+)-pinoresinol from E-coniferyl alcohol and racemic (±)-syringaresinols from E-sinapyl alcohol because of traces of contaminating oxidizing capacity. With this preparation (Table 3), a formal Km of 10 ± 6 mM and Vmax of 0.02 ± 0.02 mol s−1 mol−1 were obtained. However, with addition of fraction III, laccase, and FMN, the formal Km values (mM) were reduced to 1.6 ± 0.3, 0.100 ± 0.003, and 0.10 ± 0.01, respectively (25), whereas the Vmax values were far less affected at these concentrations of auxiliary oxidase or oxidant. These preliminary kinetic parameters are in harmony with the finding that dirigent protein does not substantially affect the rate of E-coniferyl alcohol depletion in the presence of fraction III, laccase, and FMN (Figs. 3, 4, and 6). Both sets of results are together in accord with the working hypothesis that the dirigent protein functions by capturing free-radical intermediates that then undergo stereoselective coupling.

Web Supplement

Solubilization of bound proteins was carried out at 4°C. Frozen Forsythia intermedia stems (2 kg) were pulverized in a Waring Blendor (Model CB6) in the presence of liquid nitrogen. The resulting powder was homogenized with 0.1 M KH2PO4-K2HPO4 buffer (pH 7.0, 4 liters) containing 5 mM dithiothreitol and filtered through four layers of cheesecloth. The insoluble residue was consecutively extracted, with continuous agitation at 250 rpm, as follows: with chilled (−20°C) redistilled acetone (4 liters, 3 × 30 min); 0.1 M KH2PO4-K2HPO4 buffer (pH 6.5) containing 0.1% β-mercaptoethanol (solution A, 8 liters, 30 min); solution A containing 1% Triton X100 (8 liters, 4 hours) and finally solution A (8 liters, 16 hours). Between each extraction, the residue was filtered through one layer of Miracloth (Calbiochem). Solubilization of the (+)-pinoresinol forming system was achieved by mechanically stirring the residue in solution A containing 1 M NaCl (8 liters, 4 hours). The homogenate was decanted and the resulting solution consecutively filtered through Miracloth (Calbiochem) and glass fiber (G6, Fisher Scientific). The filtrate was concentrated in an Amicon cell (Model 2000, YM 30 membrane) to a final volume of ∼800 ml and subjected to (NH4)2SO4 fractionation. Proteins precipitating between 40 and 80% saturation were recovered by centrifugation (15,000g, 30 min) and the (NH4)2SO4 pellet stored at −20°C until required.

Purification of 78-kD dirigent protein and partial purification of oxidase. The ammonium sulfate pellet (obtained from 2 kg of F. intermedia stems) was reconstituted in 40 mM MES buffer, adjusted to pH 5.0 with 6 M NaOH (solution B, 30 ml), the slurry being centrifuged (3,600g, 5 min), and the supernatant dialyzed overnight against solution B (4 liters). The dialyzed extract was filtered (0.22 μm) and the sample (35 to 40 mg proteins) was applied to a MonoS HR5/5 (50 mm by 5 mm) column equilibrated in solution B at 4°C. After eluting (flow rate 5 ml min−1 cm−2) with solution B (13 ml), proteins were desorbed with Na2SO4 in solution B by using a linear gradient from 0 to 100 mM in 8 ml and holding at this concentration for 32 ml, then implementing a series of step gradients at 133 mM for 50 ml, 166 mM for 50 ml, 200 mM for 40 ml, 233 mM for 40 ml and finally 333 mM Na2SO4 for 40 ml. Fractions capable of forming (+)-pinoresinol from E-coniferyl alcohol were eluted with 333 mM Na2SO4, combined, and stored (−80°C) until needed.

All separations on a POROS SP-M matrix (100 mm by 4.6 mm), previously equilibrated in 25 mM MES-HEPES-sodium acetate buffer (pH 5.0, solution C), were performed at a flow rate of 60 ml min−1 cm−2 and at room temperature. Fractions from 15 individual elutions from the MonoS HR5/5 column were combined (18.5 mg proteins, 180 ml) and dialyzed overnight against solution C. The dialyzed enzyme solution (190 ml) was filtered (0.22 μm) and an aliquot (47 ml) was applied to the POROS SP-M column. After elution with solution C (12 ml), the proteins were desorbed with a linear Na2SO4 gradient (0 to 0.7 M in 66.5 ml) in solution C, whereupon the concentration established was held for an additional 16.6 ml. Under these conditions, separation of four fractions (I, II, III, and IV) was achieved at ∼40, 47, 55 and 61 mS, respectively (see Fig. 2A). This purification step was repeated three times with the remaining dialyzed enzymatic extract, and fractions I, II, III, and IV from each experiment were separately combined. When protease inhibitors [that is, phenylmethanesulfonyl fluoride (0.1 mmol ml−1), EDTA (0.5 nmol ml−1), pepstatin A (1 μg ml−1), and antipain (1 μg ml−1)] were added during the solubilization and all subsequent purification stages, no differences were observed in the elution profiles of fractions I, II, III, and IV.

78-kD dirigent (from fraction I) protein. Fraction I (2.62 mg proteins, 40 ml, ∼24.6 mS) was diluted in filtered, cold distilled water until the conductivity reached ∼8 mS (final volume = 150 ml). The diluted protein solution was then applied onto a POROS SP-M column (100 mm by 4.6 mm). After elution with solution C (12 ml), fraction I was desorbed by using a linear Na2SO4 gradient from 0 to 0.25 M in 20 ml, whereupon the concentration established was held for another 25 ml. This step was followed by another linear Na2SO4 gradient from 0.25 to 0.7 M in 26 ml which was then held at 0.7 M for an additional 16.6 ml. Fractions eluted at ∼30 mS (the ionic strength of the eluent was measured with a flow-through detector) were combined (15 ml, 1.3 mg), diluted with water and rechromatographed. The resulting protein (eluted at ∼30 mS with the gradient described above) was stored (−80 °C) until needed (see Fig. 2B).

Gel filtration. An aliquot from fraction I (595.5 μg proteins, 3 ml, eluted at ∼30 mS), was concentrated to 0.6 ml (Centricon 10, Amicon) and loaded onto a S200 (73.2 cm by 1.6 cm, Pharmacia-LKB) gel chromatographic column equilibrated in 0.1 M MES-HEPES-sodium acetate buffer (pH 5.0) containing 50 mM Na2SO4 at 4 °C. An apparently homogenous 78-kD dirigent protein (242 μg) was eluted (flow rate 0.25 ml min−1 cm−2) as a single component at 133 ml (Vo = 105 ml) (Fig. 8A). Molecular weights were estimated by comparison of their elution profiles with the standard proteins, ß-amylase (200,000), alcohol dehydrogenase (150,000), bovine serum albumin (66,000), ovalbumin (45,000), carbonic anhydrase (29,000) and cytochrome c (12,400).

Polyacrylamide gel electrophoresis (PAGE) was performed in Laemmli's buffer system with gradient (4 to 15 % acrylamide, Bio-Rad) gels under denaturing and reducing conditions. Proteins were visualized by silver staining.

Oxidase from fraction III. Fraction III (2.25 mg proteins, 28 ml, ∼34 mS) was diluted exactly as above and subjected to POROS SP-M column chromatography. After elution with solution C (12 ml), proteins were desorbed as follows: after a 0.15 M Na2SO4 step of 3.3 ml, a linear Na2SO4 gradient was implemented to 0.31 M in 12.7 ml, whereupon the concentration established was held for 20 ml, then a linear Na2SO4 gradient was resumed reaching 0.39 M in 9.0 ml. This concentration was held for 20 ml, and finally a linear Na2SO4 gradient was implemented to 0.7 M in 15 ml, this being held at the final concentration for another 16.6 ml. Fractions corresponding to the peak eluted at ∼49 mS were combined (15 ml, 1.585 mg), diluted with H2O and subjected to further POROS SP-M chromatography by using a linear Na2SO4 gradient (0 to 0.7 M in 66.5 ml) in solution C. Fractions eluted between ∼48 and 57 mS were combined (0.885 mg protein) and stored (−80°C) until used without further purification (see Fig. 2C).

A 78-kD dirigent protein conferring specificity upon E-coniferyl alcohol coupling. (A) Gel filtration (S200) profile with insert showing a 27-kD subunit by SDS-PAGE analysis. (B) Ultraviolet-visible spectrum of 78-kD dirigent protein preparation.

REFERENCES AND NOTES

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
View Abstract

Stay Connected to Science

Navigate This Article