This Article Abstract Full Text (PDF) Supplemental Data All Versions of this Article: 139/3/1138    most recent pp.105.069674v1 Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via CrossRef Citing Articles via ISI Web of Science (12) Citing Articles via Google Scholar Google Scholar Articles by Gabaldón, C. Articles by Barceló, A. R. Search for Related Content PubMed PubMed Citation Articles by Gabaldón, C. Articles by Barceló, A. R. Agricola Articles by Gabaldón, C. Articles by Barceló, A. R.

BIOCHEMICAL PROCESSES AND MACROMOLECULAR STRUCTURES

Cloning and Molecular Characterization of the Basic Peroxidase Isoenzyme from Zinnia elegans, an Enzyme Involved in Lignin Biosynthesis^1,[w]

Department of Plant Biology, University of Murcia, E–30100 Murcia, Spain

	ABSTRACT

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 
The major basic peroxidase from Zinnia elegans (ZePrx) suspension cell cultures was purified and cloned, and its properties and organ expression were characterized. The ZePrx was composed of two isoforms with a M_r (determined by matrix-assisted laser-desorption ionization time of flight) of 34,700 (ZePrx34.70) and a M_r of 33,440 (ZePrx33.44). Both isoforms showed absorption maxima at 403 (Soret band), 500, and 640 nm, suggesting that both are high-spin ferric secretory class III peroxidases. M_r differences between them were due to the glycan moieties, and were confirmed from the total similarity of the N-terminal sequences (LSTTFYDTT) and by the 99.9% similarity of the tryptic fragment fingerprints obtained by reverse-phase nano-liquid chromatography. Four full-length cDNAs coding for these peroxidases were cloned. They only differ in the 5'-untranslated region. These differences probably indicate different ways in mRNA transport, stability, and regulation. According to the k_cat and apparent K_m^RH values shown by both peroxidases for the three monolignols, sinapyl alcohol was the best substrate, the endwise polymerization of sinapyl alcohol by both ZePrxs yielding highly polymerized lignins with polymerization degrees 87. Western blots using anti-ZePrx34.70 IgGs showed that ZePrx33.44 was expressed in tracheary elements, roots, and hypocotyls, while ZePrx34.70 was only expressed in roots and young hypocotyls. None of the ZePrx isoforms was significantly expressed in either leaves or cotyledons. A neighbor-joining tree constructed for the four full-length cDNAs suggests that the four putative paralogous genes encoding the four cDNAs result from duplication of a previously duplicated ancestral gene, as may be deduced from the conserved nature and conserved position of the introns.

Lignins are three-dimensional, amorphous heteropolymers that result from the oxidative coupling of three p-hydroxycinnamyl alcohols, p-coumaryl, coniferyl, and sinapyl alcohols, in a reaction mediated by both laccases and class III plant peroxidases (Ros Barceló, 1997). The cross-coupling reaction produces an optically inactive hydrophobic heteropolymer (Ralph et al., 2004) composed of p-hydroxyphenyl (H), guaiacyl (G), and syringyl (S) units, respectively.

The spatial and temporal control of lignin biosynthesis is extremely important since lignification is a metabolically costly process that requires large quantities of carbon skeletons and reducing equivalents (Amthor, 2003). Plants do not possess a mechanism to degrade lignins (Lewis and Yamamoto, 1990), so any carbon invested in lignin biosynthesis is not recoverable. Consequently, lignified cells represent a significant carbon sink (Patzlaff et al., 2003) and, as such, plants must carefully balance the synthesis of lignin polymers against the availability of resources, and this means that the monolignol biosynthetic pathway is strongly regulated (Patzlaff et al., 2003).

The metabolic flux (carbon allocation) in the phenylpropanoid pathway is controlled at multiple enzymatic levels (Boerjan et al., 2003). Studies using lignifying cell cultures of Pinus taeda, a gymnosperm, have established (Anterola et al., 1999, 2002) that both the carbon allocation to the pathway and its differential distribution into the two monolignols, p-coumaryl and coniferyl alcohol, are controlled by the rate of Phe supply and the differential modulation of cinnamate-4-hydroxylase (C4H) and p-coumarate-3-hydroxylase (C3H), respectively. In angiosperms, there is a novel branching point, the step catalyzed by coniferylaldehyde-5-hydroxylase (CAld5H), which diverts G backbones for the synthesis of S (i.e. sinapyl alcohol) moieties, the precursors of S lignins. CAld5H has not been studied in P. taeda cell cultures since they were derived from a gymnosperm, but one may extrapolate, in the absence of available data, that, like C4H and C3H in gymnosperms, CAld5H might also constitute a rate-limiting step in angiosperms. A similar role, as a rate-limiting step, has recently been proposed for the polymerization of monolignols catalyzed by basic pI class III plant peroxidases (Ipekci et al., 1999; Talas-Ogras et al., 2001; Blee et al., 2003). In this context, all the branching (also rate limiting) enzymes of the lignin biosynthetic pathway (C4H, C3H, and CAld5H) and the lignin-assembling enzyme (peroxidase) are hemoproteins. Any possible metabolic controls over these enzymes would lead to the regulation of not only the global monolignol pools in lignifying plant cells and the H/G/S ratio for carbon partitioning, but also of their rates of polymerization. Nevertheless, none of these enzymes has been cloned in Zinnia elegans, despite the frequent use of this plant as a model in lignification studies.

Z. elegans is a flowering plant belonging to the Asteraceae family. This species is commonly used as a model for studying the last step of lignin biosynthesis, i.e. the polymerization process, due to the simplicity and duality of the lignification pattern shown by stems and hypocotyls and also because of the nature of the peroxidase isoenzyme complement, which is almost completely restricted to the presence of a basic peroxidase isoenzyme (López-Serrano et al., 2004). Furthermore, Z. elegans (Fukuda, 1996) and, recently, Arabidopsis (Arabidopsis thaliana; Oda et al., 2005) offer the unique possibility of working with cell cultures that resemble differentiating xylem cells and constitute exceptionally useful models for monitoring the expression of enzymes from the lignin biosynthetic pathway, especially the segment that is concerned with the phenylpropanoid backbones (Demura et al., 2002; Milloni et al., 2002).

The lignification pattern of Z. elegans seedlings is unique in that, at a certain developmental stage, it offers simultaneously two models of lignification that closely resemble those occurring in gymnosperms and angiosperms. Thus, in 25- to 30-d-old plants, hypocotyl lignins are mainly composed of G/S units in a 42:58 ratio, while stem lignins contain significant amounts of H units in a H/G/S ratio of 22:56:22 (Ros Barceló et al., 2004). That is, S units predominate in the hypocotyl, while G units predominate in the stem. In this regard, Z. elegans hypocotyl lignins are typical of angiosperms, while the lignins of the young stem partially resemble that which occurs in gymnosperms, since (H + G) alone constitute 78% of the lignin building blocks.

Stems, hypocotyls, and transdifferentiating Z. elegans mesophyll cell cultures express the same basic peroxidase isoenzyme (López-Serrano et al., 2004). Molecular studies of this basic peroxidase isoenzyme are strongly hampered by the difficulty of obtaining the protein in large amounts (Sato et al., 1995). Transdifferentiating Z. elegans mesophyll cell cultures, leaves, stems, hypocotyls, and even roots are a limited source of the protein since it is partially covalently bound to cell walls (Masuda et al., 1983; Sato et al., 1995; Ros Barceló and Aznar-Asensio, 1999). A promising source of this isoenzyme may be Z. elegans plant cell cultures since the protein, being located in cell walls, should be rapidly secreted to the culture medium. Callus cultures derived from hypocotyls or stems grow from cambial cells, a quiescent region that also originates the vascular tissues in the shoot, including the xylem, and therefore constitutes a potential and valuable source of the protein. In fact, the Arabidopsis ATPA2 peroxidase, a plant peroxidase involved in lignification (Østergaard et al., 2000; Ehlting et al., 2005), constitutes the major peroxidase in the spent medium of Arabidopsis suspension cell cultures (SCCs) and, originally, was purified and cloned from this source (Østergaard et al., 1996).

In this article, we have purified, characterized, and cloned the major peroxidase from the spent medium of Z. elegans SCCs. The results showed that this peroxidase (ZePrx), which is expressed in Z. elegans SCCs under two distinctive forms differing in the glycosylation pattern, is coded by four full-length cDNAs differing in the 5'-untranslated regions (UTRs). These results suggest that this peroxidase isoenzyme is encoded by a complex multigene family in the genome of Z. elegans.

	RESULTS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

SCCs as a Source of the ZePrx Basic Peroxidase Isoenzymes

Isoelectric focusing (IEF) under nonequilibrium conditions was chosen to screen the presence of the ZePrx isoenzyme in the spent medium of SCCs. Results (Fig. 1) showed that transdifferentiating 3-d-old Z. elegans mesophyll cell cultures (tracheary elements [TE]), 26-d-old hypocotyls, 26-d-old stems, and SCCs expressed the same basic peroxidase isoenzyme. Quantitatively, although the level of peroxidase activity against coniferyl alcohol in SCCs was only 33% of that seen in TE when expressed on a cell basis (Table I), it was 113 times greater on a culture volume basis (Table I). This was undoubtedly due to the fact that SCCs may grow to reach cell densities of 10⁸ cells mL^–1, while TE are generally grown at cell densities of 10⁵ cells mL^–1 (Fukuda and Komamine, 1982). Cell density in transdifferentiating Z. elegans mesophyll cell cultures cannot be increased since it is critical for the differentiation process (Motose et al., 2001). Therefore, both qualitatively and quantitatively, SCC from Z. elegans constitutes a valuable source of the enzyme.

View larger version (67K): [in this window] [in a new window]   Figure 1. Basic peroxidase isoenzyme patterns obtained by IEF under nonequilibrium conditions of the spent medium fraction of transdifferentiating 3-d-old Z. elegans mesophyll cell cultures (TEs), intercellular washing fluids obtained from 26-d-old hypocotyls and stems, and the spent medium fraction from SCC. Peroxidase isoenzymes were stained with 4-methoxy--naphthol and H2O2.

The basic ZePrx isoenzyme was purified to homogeneity from the spent medium of SCC in a four-step purification protocol, which includes adsorption chromatography on phenyl Sepharose (Fig. 2A), size-exclusion chromatography on Superdex 75 (Fig. 2B), cationic exchange chromatography on SP Sepharose (Fig. 2C), and, finally, affinity chromatography on concanavalin A (Fig. 2D). This last step resolves the ZePrx isoenzyme into two isoforms: one fully glycosylated and weakly retained by the column (Fig. 2D, peak a), and the other partially glycosylated and strongly retained by the column (Fig. 2D, peak b). The M_r values estimated by SDS-PAGE for the two isoforms were 38,480 ± 1,490 and 36,470 ± 1,400, which contrast with that estimated by matrix-assisted laser-desorption ionization time-of-flight (MALDI-TOF) mass spectrometry (MS), which yielded values of 34,700 and 33,440, respectively. Table II and Figure 3 illustrate the progress of the purification of the two isoforms.

View larger version (35K): [in this window] [in a new window]   Figure 2. Purification of the basic Z. elegans peroxidase isoenzyme from the spent medium of SCCs in a four-step purification protocol, which includes adsorption chromatography on phenyl Sepharose (A), size-exclusion chromatography on Superdex75 (B), cationic exchange chromatography on SP Sepharose (C), and affinity chromatography on concanavalin A (D). This last step resolves the Z. elegans basic peroxidase isoenzyme in two isoforms: one isoform, fully glycosylated and weakly retained by the column (D, peak a), and another isoform, partially glycosylated, and strongly retained by the column (D, peak b). Profiles of peroxidase activity and protein are denoted by a continuous and broken line, respectively.

View larger version (66K): [in this window] [in a new window]   Figure 3. Protein fingerprints in SDS-PAGE at indicated steps during purification of the two isoforms of ZePrxs. From left to right, Molecular mass markers, whose values (x10–3) are drawn in the margin; crude extract protein (25 µg); protein after chromatography on phenyl Sepharose (1 µg), Superdex 75 (1 µg), and SP Sepharose (0.5 µg); and fully glycosylated isoform (0.25 µg; peak a in Fig. 2D) and partially glycosylated isoform (0.25 µg; peak b in Fig. 2D) after affinity chromatography on concanavalin A.

The two purified proteins showed absorption maxima in the visible (VIS) region at 403 (Soret band), 500, and 640 nm, which shifted to 435, 555, and 580 nm in the case of ZePrx33.44, and to 435, 560, and 580 in the case of ZePrx34.70, when the ferric enzyme was reduced by sodium dithionite. These spectral characteristics are typical and unequivocal for hemo-containing high-spin ferric secretory (class III) peroxidases (Yamazaki and Yokota, 1973).

The N-terminal amino sequences for both ZePrx33.44 and ZePrx34.70 were determined by subjecting the purified enzymes directly to Edman degradation. Microsequencing was only possible after deblocking the N terminus with pyrrolidone carboxyl peptidase, which suggests that the N terminus in both proteins is pyrrolidone carboxylic acid (Z, Pyr). After removing the N-terminal Z residue, the sequences obtained, LSTTFYDTTUPTALSTI for ZePrx33.44 (where U is undetermined) and LSTTFYDTTUUT for ZePrx34.70, suggest that both proteins share the same N terminus, being the LST motif conserved in most known class III plant peroxidase sequences (Tyson and Dhindsa, 1995).

Both ZePrx33.44 and ZePrx34.70 were subjected to deglycosylation with trifluoromethanesulfonic acid (TFMS) and further analysis of the deglycosylated proteins by SDS-PAGE. After deglycosylation, both proteins (d-ZePrx34.70 and d-ZePrx33.44) demonstrated the same mobility by SDS-PAGE (Fig. 4). The M_r value estimated by SDS-PAGE for the two deglycosylated isoforms was 33,770, which contrasts with the value of 31,460 estimated by MALDI-TOF MS. That is, the VIS properties, the N-terminal amino acid sequence, and the deglycosylation probes suggest that both ZePrx33.44 and ZePrx34.70 are the same protein, only differing in their glycosylation degree.

View larger version (59K): [in this window] [in a new window]   Figure 4. SDS-PAGE of the glycosylated isoforms ZePrx34.70 (0.5 µg) and ZePrx33.44 (0.5 µg), and the deglycosylated isoforms d-ZePrx34.70 (6.4 µg) and d-ZePrx33.44 (5.3 µg), of the basic Z. elegans peroxidase isoenzyme after deglycosylation with TFMS.

View larger version (23K): [in this window] [in a new window]   Figure 5. MALDI-TOF MS of tryptic (glyco-) peptides released from (A) the fully glycosylated isoform (ZePrx34.70) and (B) the partially glycosylated isoform (ZePrx33.44) of the basic Z. elegans peroxidase isoenzyme. Insets show in an amplified scale the tryptic fragments for both isoforms that were sequenced (arrowheads).

The peptides of m/z (M + H⁺) 1,505.8348 in ZePrx33.44 and 1,505.8392 in ZePrx34.70 were sequenced de novo from their MS³ fragment ion spectrum. This task gave the sequence MSE(I/L)GVVTGTSG(I/L)VR, which is near to the C terminus in some class III plant peroxidases (Welinder et al., 2002) and fits well with the tryptic peptide of M_r (M + H⁺) (calc) 1,505.7992. The uncertainty in the I/L residues was unresolvable at this stage since both I/Ls show isobaric mass spectra. The N-terminal amino acid sequences, including the tryptic fragments for both ZePrx33.44 (62 amino acids) and ZePrx34.70 (57 amino acids), were deposited in the Swiss-Prot database with the P84333 and P84332 accession numbers, respectively.

Cloning and Sequencing of Full-Length cDNAs Coding the ZePrx Isoenzymes

The full-length cDNAs encoding both ZePrx33.44 and ZePrx34.70 were obtained in a two-step protocol using total RNA isolated from 6-d-old Z. elegans hypocotyls. In the first step, two truncated 865-bp cDNAs were first synthesized by reverse transcription (RT)-PCR, using two (one degenerated and the other specific) primers, whose nucleotide sequences were designed to be complementary to the coding strand for the peptide sequences VRTLCGNP, present in the C-terminal sequence of both ZePrxs, and LSTTFYDT, present in the N-terminal sequence obtained for both ZePrxs.

The complete full-length sequence of the cDNAs, including the signal peptide and the 5' and 3' flanking regions, was obtained by 5'-RACE and 3'-RACE. 5'-RACE was performed using the RCPxZe-R primer while 3'-RACE PCR was performed using the RCPxZe-L primer. Both primers are shaded in green in Figure 6. Four full-length cDNAs of 1,331 bp (accession no. AJ880394 in the EMBL nucleotide sequence database; http://www.ebi.ac.uk), 1,329 bp (accession no. AJ880395), 1,304 bp (accession no. AJ880392), and 1,302 bp (accession no. AJ880393) were obtained (Fig. 6). Nucleotide sequences of the amplified full-length cDNAs were determined from both strands to ensure accuracy and were confirmed by PCR amplification of the DNA isolated from Z. elegans leaves. The four full-length cDNAs contained an identical 966-bp open reading frame (ORF). The deduced primary structure of ZePrxs is also shown in Figure 6, and contains the N-terminal amino acid sequence and the three tryptic fragments, determined experimentally. Furthermore, the uncertainty in the I/L residues of the peptide MSE(I/L)GVVTGTSG(I/L)VR, of m/z (M + H⁺) 1,505.8348 in ZePrx33.44 and 1,505.8392 in ZePrx34.70, was resolved in favor of I.

View larger version (94K): [in this window] [in a new window]   Figure 6. cDNAs encoding both ZePrx33.44 and ZePrx34.70 obtained from RNA isolated from 6-d-old Z. elegans hypocotyls and deduced amino acid sequence of ZePrxs. The complete full-length sequence of the cDNAs, including the 5' and 3' flanking regions, was obtained by 5'-RACE and 3'-RACE using the primers RCPxZe-R (5'-RACE) and RCPxZe-L (3'-RACE; shaded in green). These cDNAs isolated from Z. elegans may be categorized as full-length cDNAs from the unusual length of the 5' tails (216–245 bp), and by the fact that DNA transcription usually starts at a purine base (i.e. A), coding T (shaded in red) in the corresponding cDNAs. They also contain an identical poly(A) tract added to the 3' end, just downstream of the signal sequence 5'-AATAAA-3' (shaded in gray), which is that recognized by the poly(A) polymerase. The four full-length cDNAs contain an identical 966-bp ORF, flanked by an initiation (ATG) and a stop (TAA) codon, which translate into an identical amino acid sequence despite the observed C/T, G/A, G/C, and A/C substitutions (shaded in blue). Initiation of translation was established in the first (5' proximal) ATG codon, since it was reinforced by the presence of an A at –3 (shaded in black). The four genes codifying these cDNAs consisted of three exons and two introns. The position in which each exon was disrupted by the introns was the same in the four genes: intron of type 2 is inserted –734 bp upstream of the end codon (black arrowhead), while intron of type 3 is inserted –389 bp upstream of the end codon (white arrowhead). The mature polypeptide starts with a Q residue, which probably generates the Z residue found in the purified enzyme. The predicted polypeptide also contains a peroxidase active site signature (33-AALVIRLLFHDC, shaded in green), a peroxidase proximal hemo-ligand signature (157-EMVALSGSHTL, shaded in green), eight conserved Cys (C11, C44, C49, C87, C93, C172, C198, C287, shaded in red), which probably yield the four disulfide bridges (C11-C87, C44-C49, C93-C287, C172-C198) common in most class III plant peroxidases, two putative N-glycosylation sites (181-NSTL and 191-NRSL, shaded in yellow), and two Ca2+-binding sites (shaded in blue), one distal (D43 and D50) and the other proximal (D211 and D219), characteristics for all the active class III peroxidases. The amino acid sequence common to the root RP5a peroxidase (Sato et al., 1995) and the nucleotide base sequence common to the expressed sequence tag (AJ504423) expressed by transdifferentiating Z. elegans mesophyll cell cultures, and described by Milloni et al. (2002), are underlined. Amino acid sequences matched by tryptic fragments and the experimentally determined N-terminal sequence are double underlined.

Biochemical Properties of ZePrxs

Since glycosylation does not significantly affect the pH-dependence profile (Fig. 7A) nor the thermal stability of the ZePrxs (Fig. 7B), the oxidation of the three p-hydroxycinnamyl (p-coumaryl, coniferyl, and sinapyl) alcohols by the two ZePrxs was studied to ascertain whether glycosylation modifies the catalytic properties of the enzymes. The k_cat and apparent K_m (K_m^RH) values for the three monolignols during this peroxidase-catalyzed reaction are reported in Table IV. It can be seen that sinapyl alcohol was the best substrate for both enzymes. Although no clear tendency was found for the effect of glycosylation on the k_cat in the case of p-coumaryl and coniferyl alcohol, the full glycosylation that occurs in ZePrx34.70 reduces the k_cat for sinapyl alcohol. Likewise, the full glycosylation reduces the affinity (increases the apparent K_m^RH value) of the enzyme for both p-coumaryl and coniferyl alcohol. In the case of sinapyl alcohol, the apparent K_m^RH value was not affected significantly by glycosylation (Table IV).

View larger version (16K): [in this window] [in a new window]   Figure 7. A, pH-dependence profile. B, Temperature-dependence profile of the activity of ZePrx34.70 () and ZePrx33.44 () assayed against sinapyl alcohol. In A, peroxidase activity was measured in 50 mM Tris-acetate buffers of variable pH, containing 75 µM sinapyl alcohol and 50 µM H2O2. In B, peroxidase activity was measured in 50 mM Tris-acetate buffer, pH 5.0, containing 75 µM sinapyl alcohol and 50 µM H2O2. Values are means ± SE (n = 3).

View larger version (21K): [in this window] [in a new window]   Figure 8. GP-HPLC profiles of the products of sinapyl alcohol polymerization by ZePrx34.70 (A) and ZePrx33.44 (B), using the endwise protocol, which resolved them into a polymer of Mr 18,226 (peak 1, polymerization degree, n 87, max = 278 nm), and into oligomers with mean Mr values of 4,052 (peak 2, n 19, max = 283 nm), 1,274 (peak 3, n 6, max = 283 nm), and 667 (peak 4, n 3, max = 309 nm). Peak 5 (n 1, max = 278 nm) was unreacted sinapyl alcohol.

To study the expression of both isoforms in different organs and in the in vitro culture systems of Z. elegans, polyclonal antibodies were prepared against the fully glycosylated form of ZePrx, ZePrx34.70. Anti-ZePrx34.70 IgGs not only recognize ZePrx34.70 but also ZePrx33.44 (Fig. 9A), confirming that both peroxidases share similar epitopes.

View larger version (39K): [in this window] [in a new window]   Figure 9. A, Molecular mass markers, whose values (x10–3) are drawn in the margin, and SDS-PAGE western blots using anti-ZePrx34.70 IgGs of ZePrx34.70 (1.0 µg) and ZePrx33.44 (1.0 µg). B, SDS-PAGE western blots using anti-ZePrx34.70 IgGs of total protein (10 µg) obtained from the spent medium fraction of Z. elegans SCC before (lane a) and after (lane b) treatment of blotted membranes with sodium periodate. C, SDS-PAGE western blots using anti-ZePrx34.70 IgGs of total protein obtained from the spent medium fraction of transdifferentiating 3-d-old Z. elegans TE (10 µg protein), cotyledons (500 µg protein), and leaves (250 µg protein), after treatment of the blotted membranes with sodium periodate. D, SDS-PAGE western blots using anti-ZePrx34.70 IgGs of 3-d-old (110 µg protein) and 6-d-old (120 µg protein) roots; 3-d-old (170 µg protein), 6-d-old (80 µg protein), and 26-d-old (2.5 µg protein) hypocotyls; and 6-d-old (200 µg protein) and 26-d-old (3 µg protein) stems, after treatment of the blotted membranes with sodium periodate.

Anti-ZePrx34.70 IgGs were also tested against periodate-treated protein fingerprints obtained from TE, cotyledons, leaves, roots, hypocotyls, and stems (Fig. 9, C and D). Western blots showed that only ZePrx33.44 was expressed in TE (Fig. 9C), both ZePrx33.44 and ZePrx34.70 were expressed in roots and young hypocotyls (Fig. 9D), and only ZePrx33.44 was expressed in old hypocotyls and stems (Fig. 9D). None of the ZePrxs was significantly expressed in either cotyledons or leaves (Fig. 9C).

	DISCUSSION

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Basic ZePrxs in SCC

Class III plant peroxidases involved in lignin biosynthesis are usually classified into acidic (pI below 7.0) and basic (pI above 7.0) peroxidases. Both types of peroxidases are capable of oxidizing p-coumaryl and coniferyl alcohol. However, this situation is not so clear as regards sinapyl alcohol, which possesses a S moiety, and for which acidic peroxidases, with some exceptions (Christensen et al., 1998, 2001), are generally regarded as poor catalysts (Dean et al., 1994; Takahama, 1995; Bernards et al., 1999). This observation constitutes a central key for unraveling the specificity of lignin assembly since sinapyl alcohol is more prone to oxidation than either coniferyl alcohol or p-coumaryl alcohol (Russell et al., 1996) and suggests that, although peroxidase-catalyzed reactions are driven by redox thermodynamic forces (Folkes and Candeias, 1997), substrate accommodation in (exclusion from) the catalytic center of the enzyme determines the real role played by each peroxidase isoenzyme in lignin biosynthesis, as has been revealed from x-ray crystallographic studies (Østergaard et al., 2000).

In fact, oxidation of sinapyl alcohol by certain acidic and neutral peroxidases is sterically hindered due to unfavorable hydrophobic interactions between the sinapyl alcohol methoxy atoms and the conserved I-138 and P-139 residues at the substrate binding site of the enzyme (Østergaard et al., 2000). In the case of basic peroxidases, the capacity of these enzymes to oxidize S moieties is universally accepted (Bernards et al., 1999; Quiroga et al., 2000; Ros Barceló and Pomar, 2001; Holm et al., 2003; Sasaki et al., 2004), and this observation would explain why antisense suppression of basic peroxidases in transgenic plants produces decreased levels of both G and S lignins (Blee et al., 2003), while antisense suppression of acidic peroxidases produces only decreased levels of G lignins (Li et al., 2003).

In accordance with their key role in lignin biosynthesis, cationic (basic) peroxidases are differentially expressed during the transdifferentiation of Z. elegans mesophyll cell cultures, where they act as molecular markers of xylogenesis (Masuda et al., 1983; Sato et al., 1995; López-Serrano et al., 2004). These same peroxidase isoenzymes may also be recovered in the spent medium protein fraction from Z. elegans SSC (Fig. 1) and show a special ability for oxidizing sinapyl alcohol when compared with other putative peroxidase substrates (Table III).

Heterogeneity of ZePrxs

The basic peroxidase isoenzyme isolated from Z. elegans shows homogeneity as regards the pI under IEF (Fig. 1), but can be resolved by chromatography on concanavalin A (Fig. 2D) in two isoforms: one partially glycosylated (M_r 33,440) and strongly retained by the column, and another fully glycosylated (M_r 34,700) and weakly retained by the column. This heterogeneity has been described for some basic peroxidases (Rasmussen et al., 1991; Wan et al., 1994) and it is not surprising since full glycosylation, which leads to the formation of complex glycans in plant glycoproteins, initially involves the addition of Fuc and Xyl end moieties to the Man-rich core and, finally, substitution of the Man end moieties by Gal units (Wilson, 2002). Undoubtedly, the coating of the Man-rich core with Fuc, Xyl, or Gal end units reduces the strength with which the glycoprotein will be retained by concanavalin A, and this allows both glycoproteins to be separated.

In any case, deglycosylation with TFMS of both ZePrx33.44 and ZePrx34.70 yielded the same deglycosylated protein (Fig. 4) with an estimated M_r (MALDI-TOF) of 31,460. That is, the heterogeneity of the Z. elegans peroxidases may be attributed to the heterogeneity of the glycan moieties, as has previously been reported for peroxidases from flax (Gaudreault and Tyson, 1988), barley (Rasmussen et al., 1991; Johansson et al., 1992), and peanut (Wan et al., 1994). This assumption was further strengthened by the total similarity of the N-terminal sequence determined for both peroxidases (LSTTFYDTT), and by the 99.9% similarity of the tryptic fragment fingerprints obtained by RP nano-LC and MALDI-TOF MS (Fig. 5).

Cloning and Primary Structure of ZePrxs

From the N-terminal amino acid sequence and a C-terminal tryptic fragment, four full-length cDNAs encoding the ZePrxs were isolated (Fig. 6). The four full-length cDNAs contain an identical 966-bp ORF encoding 321 amino acids, which coded for the primary structure of ZePrxs shown in Figure 6. That these cDNAs isolated from Z. elegans may be categorized as full-length cDNAs may be deduced from the unusually great length of the 5' tails (216–245 bp), and by the fact that DNA transcription usually starts at a purine base (i.e. adenine [A]; Nishiyama et al., 2003), coding thymine (T) in the corresponding cDNAs, as was the case here (Fig. 6).

The four full-length cDNAs contain (1) a nonconserved 5'-UTR of variable length (245 bp in AJ880394, 243 bp in AJ880395, 218 bp in AJ880392, and 216 bp in AJ880393); (2) a 966-bp ORF, whose nucleotide sequences translate into an identical amino acid sequence, despite the observed C/T, G/A, G/C, and A/C substitutions (Fig. 6); (3) an identical 120-bp 3'-UTR; and (4) an identical polyadenylic acid [poly(A)] tract added to the 3' end, just downstream of the signal sequence 5'-AATAAA-3' (Fig. 6; 5'-AAUAAA-3' in the mRNAs), which is recognized by the poly(A) polymerase. Although little is known about the precise function of the 5'-UTRs, it is accepted (Kozak, 1991) that these regions contain sequences that can form secondary structures and interact with proteins that regulate transport, translation, and stability of the mRNAs. Thus, the differences observed in the 5'-UTRs for the four full-length mRNAs probably indicate different ways in their transport, stability, and regulation.

Since translation of most eukaryotic mRNAs starts at the first (5' proximal) AUG codon, in accordance with the scanning behavior of the ribosomal 40S unit (Kozak, 1991), the start codon for the four ORFs was established at 246 bp in AJ880394, 244 bp in AJ880395, 219 bp in AJ880392, and 217 bp in AJ880393, downstream of the 5' end. None of these cDNAs contains the start codon preference found in dicots (5'-AAAaugGC-3'), which would challenge the above rule. Besides, this putative position for the start codon is reinforced by the presence of an A at –3 (Fig. 6), which is known to lend translational weight to the first recognizable 5' proximal AUG codon (Kozak, 1991), as has been shown in Arabidopsis (Østergaard et al., 1998).

The position of the start codon for the four ORFs described in Figure 6 suggests that the immature polypeptide contains a signal peptide (N-terminal propeptide) of 30 amino acids (MSYHKSSGTTLMVPLFMLLISVNYFMSCNA), which directs the polypeptide chain to the endoplasmic reticulum membrane. In the absence of a hydropathy profile (Nielsen et al., 1997), the signal peptide cleavage site was predicted between 30A and 31Q (A||QLS), in accordance with the usual cleavage sites described for eukaryotes (Nielsen et al., 1997). Furthermore, this position was confirmed experimentally from the N-terminal sequence obtained for both ZePrxs (QLS). The putative N-terminal propeptide was characterized by the presence of a positive charged n-region, MSYHK, a hydrophobic h-region with Leu-rich zips, LMVPLFMLLI, and a neutral-polar c-region, SCN. The proposed cleavage site also fulfills the (–3,–1) rule (Nielsen et al., 1997), which states that the residues at positions –3 (C) and –1 (A) (relative to the cleavage site) must be small and neutral. Establishment of the polypeptide cleavage site in this position also suggests that, unlike certain Arabidopsis peroxidases of an unknown function (Welinder et al., 2002), the ZePrxs do not contain an N-terminal extension (an NX propeptide according to Welinder's nomenclature).

Like other class III plant peroxidases, the mature polypeptide (Fig. 6) starts with a Q residue, which probably generates the Z residue found in the purified enzyme, producing a mass loss of 17 D. This is a well-documented modification that can occur during protein handling (Khandke et al., 1989), although enzymes that convert glutaminyl peptides into pyroglutamyl peptides have been reported in mammals, where they participate in the biosynthesis of many hormones and neurotransmitter peptides (Busby et al., 1987).

Deduction of the signal peptide M_r resulted in a mature polypeptide M_r of 30,863, which, with the additional contribution of hemo b (616), yields a putative mature protein of M_r 31,479. The value fits well with the M_r obtained for the deglycosylated forms of both ZePrx (31,460 – Z + Q = 31,477). The mature protein is of a basic nature and has a theoretical pI of 8.47. In the case of peroxidases, predicted pIs between 5 and 10 are generally two units lower than the experimental values because the two calcium ions and the hemo are not included in the calculation (Welinder et al., 2002). After this correction, the theoretical pI for both ZePrxs (8.47 + approximately 2.0) fits well with the pI value of about 10.2 to 10.5 determined experimentally (López-Serrano et al., 2004).

The mature polypeptide also contains the following peptide motif around the eighth C (C287), TGTSGIVRTLCGNPS·, whose partial sequence, TGTSGIVR, was also identified in the tryptic fragment of m/z 1,505.8348 in ZePrx 33.44 and 1,505.8392 in ZePrx34.70. The proximity (four amino acids) of the eighth C to the end codon suggests that this polypeptide has no C-terminal propeptide extension to target the proteins for vacuolar transport, which, when present in class III plant peroxidases (Welinder et al., 2002), usually adds a 20- to 30-amino acid tail downstream of the eighth C. Since the polypeptide does not contain any of the endoplasmic reticulum motifs for retrograde transport (the KDEL, HDEL, and DEL tails), it can be affirmed that the protein follows the default pathway, which ends in the cell wall. This cell wall localization of the protein is corroborated from the recovery of the enzyme in the spent medium fraction of Z. elegans SCC.

Possible Physiological Roles of ZePrxs

The best way to ascertain for which particular metabolic reaction of the complex process of lignin assembly ZePrxs have been designed in the course of vascular plant evolution is to compare the apparent K_m^RH values for the three monolignols. An example of the importance of these values is the observation that microsomal 5-hydroxylases, previously named ferulate-5-hydroxylases, hydroxylate coniferyl aldehyde (K_m = 1 to 2.7 µM), and coniferyl alcohol (K_m = 3 µM) with a greater affinity than ferulate (K_m = 286 to 1,000 µM; Humphreys et al., 1999; Osakabe et al., 1999), leading to the tendency to call them as coniferyl aldehyde-5-hydroxylases (CAld5H).

To cast light on this question, the apparent K_m values of the ZePrxs for the three monolignols (K_m^RH) during the peroxidase reaction were examined in detail. From these data (Table IV), it can be concluded that ZePrxs show the lowest apparent K_m^RH values (in the micromolar range) for sinapyl alcohol, illustrating the great affinity of ZePrxs for this monomeric lignin precursor. Besides, the full glycosylation occurring in ZePrx34.70 reduces the affinity (increases the apparent K_m^RH value) of the enzyme for both p-coumaryl and coniferyl alcohol.

The apparent K_m^RH values for sinapyl alcohol shown by the ZePrxs were in the same range as those shown by the peroxidase-preceding enzymes of the lignin biosynthetic pathway, cinnamyl alcohol dehydrogenase and microsomal 5-hydroxylases, which also use both cinnamyl alcohols and aldehydes as substrates (Humphreys et al., 1999; Osakabe et al., 1999). The micromolar nature of the K_m^RH values for these enzymes implicitly indicates that the one-way highway of lignin macromolecule construction, which properly starts with the reduction of cinnamoyl-CoAs by cinnamoyl-CoA reductases, contains no metabolic potholes in which the lignin building blocks might be accumulated. This observation is in accordance with the predictions that may be made from the observed pools and metabolic fluxes of monolignols in lignifying cells (Anterola et al., 1999).

The overall kinetic properties during the peroxidase reaction of these hemo proteins suggested again that sinapyl alcohol is the better substrate for both ZePrxs, since it shows not only the lowest apparent K_m^RH values but also the highest k_cat, which leads to the highest catalytic efficacy (k_cat/K_m^RH; Table IV). This is in accordance with the oxido/reduction potentials for the three monolignols since sinapyl alcohol is more prone to oxidation than coniferyl alcohol and much more so than p-coumaryl alcohol, which is the least reactive substrate (Russell et al., 1996). Furthermore, the reactivity of ZePrxs for synapyl alcohol is such that sinapyl alcohol oxidation by both ZePrxs yields oxidation products of an oligomeric nature, with M_r 18,226 and polymerization degrees, n 87 (Fig. 8), which supports a role for these peroxidases in lignin biosynthesis. This observation led us to conclude that ZePrxs, unlike Arabidopsis ATP2 and HRP A2s (Østergaard et al., 2000; Nielsen et al., 2001), have no steric hindrance in the hemo crevice for actively discriminating among the three p-hydroxycinnamyl alcohols, and suggests a high degree of metabolic plasticity (versatility) for this basic peroxidase, making it one of the possible driving forces in the evolution of plant lignin heterogeneity.

We should mention, at this point, that the versatility of certain enzymes is one of the main driving forces in the evolution of land plants, and that there is a general consensus (Boerjan et al., 2003) that such enzymes confer high metabolic plasticity to the lignin biosynthetic pathway. The metabolic plasticity of this basic peroxidase, capable of catalyzing with high affinity and high catalytic activity the polymerization of the three p-hydroxycinnamyl alcohols, means that the heterogeneity of lignin precursors is not prohibitive, nor is its metabolic cost terribly expensive for the plant. In fact, the versatility of this basic peroxidase confers a certain sense to the heterogeneity (three p-hydroxycinnamyl alcohols) of the lignin biosynthetic pathway, as far as we know in angiosperms.

The cDNAs Coding for ZePrxs Are Both Expressed and Translated in Z. elegans TE

ZePrxs that differ in the glycosylation pattern not only show differential kinetic properties but also a differential organ expression when analyzed by western blot. Thus, when anti-ZePrx34.70 IgGs, which recognize both ZePrx33.44 and ZePrx34.70 (Fig. 9A), were tested against protein fingerprints obtained from SCC, TE, cotyledons, leaves, roots, hypocotyls, and stems, it was found that ZePrx33.44 was expressed in SCC, TE, roots, hypocotyls, and stems, while ZePrx34.70 was only expressed in SCC, roots, and young hypocotyls (Fig. 9, B–D).

The immunological reactivity of the root Z. elegans peroxidases against anti- ZePrx34.70 IgGs (Fig. 9D) is supported by the observation that a BrCN-derived peptide sequence, VALSGUHTUG (where U is undetermined), which is contained in the Z. elegans peroxidase, RP5a, purified from roots, and considered as a marker of xylogenesis in Z. elegans (Sato et al., 1995), is self-contained in the primary amino acid sequence of ZePrxs, concretely in the peptide 159-VALSGSHTLG (Fig. 6). These results suggest that both peroxidases are similar, if not the same.

Likewise, the immunological reactivity of the TE Z. elegans peroxidases against anti-ZePrx34.70 IgGs (Fig. 9C) is in accordance with the observation that an expressed sequence tag isolated from transdifferentiating Z. elegans mesophyll cells (AJ504423) by Milloni et al. (2002) is shared with the four full-length cDNA clones isolated from Z. elegans hypocotyls, the region comprised between –4 to –221 bp upstream the end codon (TAA; Fig. 6, underlined). These results suggest that the cDNAs coding for ZePrxs are both expressed and translated in Z. elegans TE.

ZePrxs Are Encoded by a Complex Multigene Family in the Genome of Z. elegans

A neighbor-joining tree (Kumar et al., 2001) constructed for the four full-length cDNAs isolated from hypocotyls suggests that the four putative paralogous genes encoding the four cDNAs could arise from the duplication of a previously duplicated ancestral gene, since AJ880394 and AJ880395, on the one hand, and AJ880392 and AJ880393, on the other hand, are clustered together. Preliminary results obtained in our laboratory suggest that the four genes codifying ZePrxs consisted of three exons and two introns, one of type 2 and the other of type 3, according to Tognolli et al.'s (2002) nomenclature. The nature and position (see Fig. 6) of these introns suggest that ZePrxs belong to the peroxidase gene subgroup of class a, already described in Oryza (Passardi et al., 2004a) and Arabidopsis (Tognolli et al., 2002). Interestingly, the position in which each exon is disrupted by the introns was the same in the four genes (Fig. 6, arrowheads, –389 bp upstream the end codon for intron 3 and –734 bp upstream the end codon for intron 2). Furthermore, the base pair sequence of intron 2 was totally conserved in the genes codifying AJ880393 and AJ880395 (accession no. AJ971430), and in the genes codifying AJ880392 and AJ880394 (accession no. AJ971431). Finally, search homologies suggest that both introns, AJ971430 and AJ971431, are paralogous, showing 94.7% similarity. These preliminary results suggest that duplications of the ZePrxs occurred by recombination within introns, as has been described for two HRP basic genes coded in tandem (Fujiyama et al., 1988).

In Arabidopsis, analysis of the genome sequence revealed the existence of pairs of chromosomal blocks that exhibit high-level synteny and a similar gene content (Arabidopsis Genome Initiative, 2000). These blocks span almost the whole of the genome without overlapping and are thought to be remnants from a recent genome-wide duplication. Statistical and phylogenetic studies suggest that these ancient duplication events occurred before the divergence of Arabidopsis from other dicots and may constitute an even earlier event predating the monocot-dicot divergence (Kim et al., 2005). Although it has been reported that, after a duplication event, degenerative mutations will probably eliminate many duplicated genes from the genome (Lynch and Conery, 2000), the remaining duplicated genes are thought to be a source of biochemical diversity preserved by subfunctionalization (Lynch and Force, 2000).

Neofunctionalization (acquisition of a novel function in a protein owing to mutations or duplications at the DNA level during evolution) and subfunctionalization (acquisition of a new expression profile for a duplicated gene) are not new facts when we treat genes codifying class III plant peroxidases. In fact, class III plant peroxidases belong to a multigene family, whose evolution seems to be correlated with the increasing complexity of plant cell wall architecture and the diversification of their biotopes and pathogens (Duroux and Welinder, 2003; Passardi et al., 2004b). To exemplify this, a high duplication rate in some plants has led to large multigene families, as suggested by the presence of 138 peroxidase-encoding genes in Oryza (Passardi et al., 2004a) and of 73 peroxidase-encoding genes in Arabidopsis (Tognolli et al., 2002; Welinder et al., 2002; Valério et al., 2004). Furthermore, in Arabidopsis, two identical peroxidase genes are represented by AtP11 (AtP11.1 and AtP11.2; Welinder et al., 2002). Either subfunctionalization or neofunctionalization could explain the conservation of these duplicate genes and the presence of such large multigene families in which each paralog (homologous genes produced by duplication within a genome) could become specialized for a determined task.

At this stage, it is not easy to trace a forward relationship (mRNAs versus proteins) between the four full-length cDNAs and the two ZePrxs, since the four full-length cDNAs code for an identical protein primary structure and the two ZePrxs only differ in their glycosylation pattern. However, since both ZePrxs differ in their organ expression (Fig. 9) and catalytic properties (Table IV), it is probable that we are dealing here with two new and very particular examples of subfunctionalization and neofunctionalization performed at the posttranslational level. Subfunctionalization and neofunctionalization are not new facts when talking of enzymes of the lignin biosynthesis pathway. In fact, retained copies of Phe-ammonia lyase and cytochrome P450 monooxygenases in plant genomes led to an increased gene dosage and provided a fitness advantage since some of them acquired new functions through sequence divergence (Kim et al., 2005). The cytochrome P450 monooxygenase, CAld5H, constitutes the best example, since it appears as the main driving force in the gymnosperm/angiosperm transition of plant lignin heterogeneity.

	MATERIALS AND METHODS

TOP ABSTRACT RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

 

Cell Cultures and Protein Fractions

Zinnia elegans (cv Envy, Chiltern Seeds, Cumbria, England) were germinated in a petri dish for 6 d in distilled water and the seedlings were sterilized in a 70% (v/v) aqueous solution of ethanol for 4 min, followed by 10% (v/v) aqueous solution of Domestos for 40 min. After three washes in sterile water, 1.0-cm hypocotyl explants were placed on Murashige and Skoog medium, supplemented with 250 mg L^–1 casein hydrolysate, 25 g L^–1 Suc, 1 mg L^–1 calcium pantotenate, 100 mg L^–1 myoinositol, 0.01 mg L^–1 biotine, 3 mg L^–1 nicotinic acid, 2 mg L^–1 thiamine and 1 mg L^–1 piridoxine, 1 mg L^–1 1-naphthalenacetic acid, and 1 mg L^–1 kinetin. Cultures were grown in the dark at 25°C and callus developed in 25 to 30 d. Friable calluses were subcultured for 2 years. Suspension cell cultures were established from friable calluses in 250-mL flasks. Cultures were grown on a rotator shaker (100 rpm) at 25°C and in darkness in 90 mL of culture medium, and subcultured every 8 d by diluting 1:1 (v/v) in fresh medium. Cell growth was stabilized by a 4-month period of repetitive culture. After this time, SCCs were grown until they reached 68.70% (±10.99) packed cell volume, corresponding to a cell density of 1.37 (±0.45) 10⁸ cells mL^–1.

Transdifferentiating Z. elegans mesophyll cell cultures were established from true leaves from 14-d-old seedlings, which were surface sterilized in 5% (v/v) commercial NaOCl and rinsed in sterile distilled water. Single cells were isolated and cultured for 3 d in a differentiating medium as described (Fukuda and Komamine, 1982; López-Serrano et al., 2004).

In both cases, cells were separated from the culture medium by centrifugation at 100g for 1 min at 4°C, the supernatant constituting the spent medium (apoplastic) protein fraction. This protein fraction was desalted by chromatography on PD-10 Sephadex G-25 (Amersham Bioscience) columns equilibrated in 50 mM sodium acetate buffer, pH 5.0, containing the protease inhibitors 1.0 mM phenylmethanesulfonyl fluoride (PMSF) and 1.0 mM benzamidine, and concentrated using Ultrafree-0.5 (Millipore).

Other Plant Materials and Protein Fractions

Seedlings of Z. elegans were also grown for 3, 6, and 26 d in a greenhouse under daylight conditions at 25°C as described (Ros Barceló and Aznar-Asensio, 1999). Total protein was extracted from cotyledons, roots, hypocotyls, stems, and leaves by homogenizing in 25 mM MES buffer, pH 6.0, containing 1.0 mM EDTA, 1.0 mM dithiothreitol (DTT), and 1.0 mM PMSF, and centrifugation at 18,670g for 20 min at 4°C. Desalting and protein concentration of the supernatant was performed with Amicon Ultra-4 (Millipore).

To recover the intercellular washing fluids, sections (5.0 g fresh weight) of 26-d-old hypocotyls (or stems), measuring less than 5 mm, were soaked in deionized water and subsequently vacuum infiltrated for 5 min at 1.0 kPa and 4°C with 50 mM Tris-acetate buffer, pH 5.0, containing 1.0 M KCl and 50 mM CaCl₂. The sections were quickly dried and centrifuged in a 25-mL syringe barrel placed within a centrifuge tube at 900g for 5 min at 4°C. Protein fractions were desalted by chromatography on PD-10 Sephadex G-25 (Amersham Bioscience) columns equilibrated in 50 mM sodium acetate buffer, pH 5.0, containing the protease inhibitors 1.0 mM PMSF and 1.0 mM benzamidine, and concentrated using Ultrafree-0.5 (Millipore).

Purification of ZePrxs

To purify ZePrxs, 18.50 L of spent medium was salted out with (NH₄)₂ SO₄ up to 95% saturation, and the total spent medium protein was recovered by centrifugation at 30,000g for 20 min at 4°C. The pellet was resuspended in 50 mM Tris-HCl, pH 7.5, and dialyzed in this buffer overnight. The dialyzed sample was concentrated in Centricon Ultra, and dissolved in 1.5 M (NH₄)₂ SO₄, this fraction constituting the initial (crude) protein fraction from which spent medium ZePrxs were purified by five-step chromatography at 5°C, using the Bio-Rad Econo system (Bio-Rad Laboratories).

In the first step, the crude protein was chromatographed on a phenyl Sepharose 6 Fast Flow (Amersham Biosciences) 42.5 x 1.0-cm gel-bed column at a flow rate 1.0 mL min^–1, and fractions of 5.0 mL were recovered. The eluent chromatography program was as follows: from 0 to 115 min (100% A, 0% B), from 115 to 200 min (0%–100% B), and from 200 to 300 min (100% B), where buffer A was 50 mM Tris-HCl, pH 7.5, containing 1.5 M (NH₄)₂ SO₄, and buffer B was 50 mM Tris-HCl, pH 7.5.

The second step involved size exclusion chromatography on Superdex 75 (Amersham Biosciences). For this, the peroxidase-rich fractions obtained from the phenyl Sepharose 6 Fast Flow chromatography were dialyzed against 50 mM Tris-HCl, pH 7.5, and chromatographed on a Superdex 75 53.5 x 1.6-mL gel-bed column equilibrated with 50 mM Tris-HCl, pH 7.5. The flow rate was 0.5 mL min^–1 and fractions of 2.5 mL were recovered.

The third step involved ion-exchange chromatography on SP Sepharose Fast Flow (Amersham Biosciences). For this, the peroxidase-rich fractions obtained from the size exclusion chromatography were dialyzed against 50 mM 3-[cyclohexylamino]-1-propane-sulfonic acid (CAPS), pH 9.5, and loaded on 44 x 1.0-cm gel-bed column equilibrated with 50 mM CAPS, pH 9.5, at a flow rate of 1.0 mL min^–1. Fractions of 1.0 mL were recovered. The eluent chromatography program was as follows: from 0 to 35 min (100% A, 0% B), from 35 to 100 min (0%–100% B), and from 100 to 200 min (100% B), where buffer A was 50 mM CAPS, pH 9.5, and buffer B was 50 mM CAPS, pH 11.5.

The fourth step involved affinity chromatography on concanavalin A-Sepharose 4B (Sigma). For this, the peroxidase-rich fractions obtained from the ion-exchange chromatography were dialyzed against 50 mM Tris-HCl, pH