Molecular Plant Advance Access published online on July 14, 2009
Molecular Plant, doi:10.1093/mp/ssp044
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Feruloylated Arabinoxylans Are Oxidatively Cross-Linked by Extracellular Maize Peroxidase but Not by Horseradish Peroxidase
a The Edinburgh Cell Wall Group, Institute of Molecular Plant Sciences, School of Biological Sciences, The University of Edinburgh, Daniel Rutherford Building, The King's Buildings, Edinburgh EH9 3JH, UK
b Present address: Ecologie Microbienne/UMR CNRS 5557 USC INRA 1193, Université Claude Bernard—Lyon 1, Bâtiment Gregor Mendel, 16 rue Dubois, F-69622 Villeurbanne Cedex, France
1 To whom correspondence should be addressed. E-mail S.Fry{at}ed.ac.uk, fax +44 131 650 5392, tel. +44 131 650 5320.
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
Covalent cross-linking of soluble extracellular arabinoxylans in living maize cultures, which models the cross-linking of wall-bound arabinoxylans, is due to oxidation of feruloyl esters to oligoferuloyl esters and ethers. The oxidizing system responsible could be H2O2/peroxidase, O2/laccase, or reactive oxygen species acting non-enzymically. To distinguish these possibilities, we studied arabinoxylan cross-linking in vivo and in vitro. In living cultures, exogenous, soluble, extracellular, feruloylated [pentosyl-3H]arabinoxylans underwent cross-linking, beginning abruptly 8 d after sub-culture. Cross-linking was suppressed by iodide, an H2O2 scavenger, indicating dependence on endogenous H2O2. However, exogenous H2O2 did not cause precocious cross-linking, despite the constant presence of endogenous peroxidases, suggesting that younger cultures contained natural cross-linking inhibitors. Dialysed culture-filtrates cross-linked [3H]arabinoxylans in vitro only if H2O2 was also added, indicating a peroxidase requirement. This cross-linking was highly ionic-strength-dependent. The peroxidases responsible were heat-labile, although relatively heat-stable peroxidases (assayed on o-dianisidine) were also present. Surprisingly, added horseradish peroxidase, even after heat-denaturation, blocked the arabinoxylan-cross-linking action of maize peroxidases, suggesting that the horseradish protein was a competing substrate for [3H]arabinoxylan coupling. In conclusion, we show for the first time that cross-linking of extracellular arabinoxylan in living maize cultures is an action of apoplastic peroxidases, some of whose unusual properties we report.
Key Words: Cell wall • cross-links • phenolics • ferulate • peroxidase • soluble extracellular polysaccharides • Zea mays L
Received for publication March 4, 2009. Accepted for publication June 8, 2009.
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
Primary cell walls in the Poaceae (grasses and cereals) and their close relatives possess certain unique features, being low in pectins (
5%), high in arabinoxylans (30%), and relatively high in non-lignin phenolics (0.5–1.5% dry weight) (Harris et al., 1997; Fry, 2000; Fincher, 2009). One specific phenolic compound, ferulic acid, is attached to Poaceae arabinoxylans as a side-chain, and throughout this paper, the term ‘arabinoxylan’ implies ‘feruloylated arabinoxylan’, unless otherwise stated. It has been speculated that feruloyl ester groups on adjacent arabinoxylan chains become oxidatively coupled during cell-wall development in vivo, thus forming inter-polymeric covalent bonds that cross-link these chains and tighten the cell wall, and that consequently feruloyl-bridged arabinoxylans perform a role in the Poaceae similar to that of calcium- or borate-bridged pectins in plants generally (Hatfield et al., 1999; Pérez et al., 2003; O'Neill et al., 2004; Proseus and Boyer, 2008). In studies of the biology of polysaccharide cross-linking, it is essential to distinguish the cross-linking reactions that may occur in vivo from those that can readily be demonstrated to occur under artificial conditions in vitro. Inter-polymeric cross-linking of feruloyl-polysaccharides can be achieved in vitro. For example, galactomannans, artificially esterified with ferulic acid, have been shown to gel when exposed to oxidizing agents, and subsequently to release diferulates (also known as dehydrodiferulates) when saponified (Geissmann and Neukom, 1971). Feruloylated pectins can likewise be cross-linked in vitro (Rombouts and Thibault, 1986; Guillon and Thibault, 1990; Micard and Thibault, 1999), and feruloylated arabinoxylans isolated from wheat flour and wheat bran have been shown to behave in a similar way on treatment with peroxidase + H2O2 (Schooneveld-Bergmans et al., 1999).
Various isomers of diferulic acid have been obtained by alkali hydrolysis of cell walls and from in-vitro cross-linked feruloyl-polysaccharides (Markwalder and Neukom, 1976; Harris and Hartley, 1980; Hartley and Harris, 1981; Ralph et al., 1994; Brett et al., 1999; Waldron et al., 1996; Bunzel et al., 2004, 2005), and may contribute to the natural cross-linking of arabinoxylans in vivo. The diferulate:ferulate ratio was shown to increase when isolated maize cell wall preparations containing bound peroxidase were exposed in vitro to dilute hydrogen peroxide (Grabber et al., 1995).
There is some evidence that the diferulates released by alkali from cell walls derive from feruloyl arabinoxylans: for example, a 5,5'-di-(feruloyl-arabinosyl-xylobiose) was isolated from maize bran after Driselase digestion (Ishii, 1991). Furthermore, 5,5'-di-(feruloyl-arabinose) (Saulnier et al., 1999) and 8-O-4-di-(feruloyl arabinose) (Allerdings et al., 2005) have been isolated after partial acid hydrolysis of maize bran arabinoxylans. As yet, no oligosaccharides linked to other diferulate isomers, or to larger oxidation products, have been fully characterized.
Although diferulates may contribute to polysaccharide cross-linking, ferulate-products larger than dimers (oligoferulates) often predominate in vivo (Fry, 1984; Fry et al., 2000) and in vitro (Ward et al., 2001). As well as their being quantitatively predominant, it is also more likely that the larger coupling products could act as real ( = inter-polysaccharide) cross-links rather than forming intra-polysaccharide loops* (Fry, 2004a). Therefore, assays of dimers alone may seriously underestimate the degree of feruloyl-polysaccharide cross-linking (if oligoferulates predominate) or overestimate it (if diferulates form mainly intra-polysaccharide loops). Certain specific isomers of triferulic acid (Bunzel et al., 2004, 2005, 2006; Funk et al., 2005; Rouau et al., 2003) and tetraferulic acid (Bunzel et al., 2006) have been isolated and structurally elucidated from primary cell walls in grass and cereal species, particularly from maize bran.
A more general approach for detecting, as a group, total coupling products including those larger than dimers is to feed the cells with [14C]cinnamate, a precursor of ferulate; all ferulate-derived substances can then be detected quantitatively, thanks to their radioactivity, whether or not their physicochemical properties (susceptibility to hydrolysis, solubility, chromatographic behaviour, etc.) could have been predicted (Fry, 1984; Fry et al., 2000; Lindsay and Fry, 2008; Burr and Fry, 2009). In maize cell cultures fed with [14C]cinnamate, inhibition of peroxidase action by treatment with iodide, dithiothreitol (DTT), or cysteine caused an increased proportion of [14C]feruloyl-arabinoxylans to be released into the medium, supporting the hypothesis that peroxidase action is responsible for polysaccharide cross-linking in the cell wall in vivo (Lindsay and Fry, 2008).
During the progression of maize cell cultures through the culture cycle, the site of diferulate production shifts from the protoplasm to the cell wall (Fry et al., 2000). In younger cultures, pre-cross-linked feruloyl-arabinoxylans are proposed to be secreted into the apoplast as large coagula, which would hydrogen-bond only loosely to the cellulose microfibrils already present in the cell wall, thus maintaining a high extensibility in the growing cell wall. In older cultures, in contrast, the feruloyl-arabinoxylans are secreted into the cell wall as individual, non-cross-linked polysaccharide chains, which would not only form extensive hydrogen bonds with the cellulose, but would also be capable of subsequently forming inter-arabinoxylan covalent bonds via diferulates and oligoferulates, consequently clamping the microfibrils and transforming the cell wall into a less extensible network.
If oxidative coupling occurs between feruloyl groups on adjacent polysaccharide chains in vivo, producing a cross-linked network, the effective relative molecular mass (Mr) of the polysaccharide is expected to increase considerably. However, the cell wall polysaccharides, being already hydrogen-bonded into a dense network, are insoluble, and therefore recalcitrant to Mr measurements. Solubilization of wall-bound arabinoxylans is only possible by (pre-) treatments, such as with aqueous alkali or hydroxylamine (Mares and Stone, 1973) or methanolic sodium methoxide (Morrison, 1977), which are likely to cleave the ester-based cross-links under investigation. Fortunately, maize cell-suspension cultures slough some of their wall-related polysaccharides into the culture medium, where, since they remain soluble, they provide a convenient model system for studying polysaccharide cross-linking in vivo. Indeed, when cultured maize cells were fed [3H]arabinose, the subsequently released radiolabelled soluble extracellular polysaccharides (mainly feruloylated arabinoxylans) were shown to undergo an abrupt size increase from 1 to >17 MDa at a time-point between 8 and 11 d into the culture cycle (Kerr and Fry, 2003). This increase was delayed by exogenous sinapic acid or chlorogenic acid, both of which are low-Mr substrates competing with the feruloyl residues of arabinoxylans to be oxidatively coupled (Kerr and Fry, 2004). Consequently, the feruloyl side-chains of arabinoxylans are implicated in the Mr increase, and therefore in polysaccharide cross-linking.
An even further simplified experimental system for studying ferulate-based cross-linking in vivo involves the use of an exogenous radiolabelled feruloyl oligosaccharide, 5-O-feruloyl-
-L-arabinofuranosyl-(1
3)-β-D-xylopyranosyl-(14)-D-xylose (‘FAXX’, now referred to as A5f3X according to the abbreviated nomenclature of Fauré et al., 2009). This compound chemically resembles a feruloyl-arabinoxylan but physically differs from it in being small enough to permeate the cell wall matrix far more effectively than an exogenous feruloyl-polysaccharide could. Exogenous A5f3X was shown to become covalently bonded, via its feruloyl moiety, to cell-wall polysaccharides in living maize cell-suspension cultures. The bonding was to already wall-bound polysaccharides in the apoplast, not to nascent polysaccharides in the endo-membrane system. When [feruloyl-14C]A5f3X was used as the model compound, radiolabelled diferulate and larger coupling products were generated (Encina and Fry, 2005). Studies with this experimental system also revealed the presence in the apoplast of a heat-labile, low-Mr inhibitor of feruloyl coupling at specific developmental stages, which is relevant to the physiological control of the cross-linking process (Encina and Fry, 2005).
Finally, Burr and Fry (2009) demonstrated that the increase in Mr of the soluble extracellular arabinoxylans occurs simultaneously with the consumption of arabinoxylan-bound ferulic acid and the production of oligoferulates. This ties in with results of Fry et al. (2000), which suggest that oligoferulates play a greater role in polysaccharide cross-linking than diferulates, and also with results of Schooneveld-Bergmans et al. (1999), who showed that in-vitro oxidative cross-linking of feruloyl arabinoxylans from wheat led to a decrease in the total of measurable ferulate products, which consisted at that time solely of ferulate and diferulates.
Previous work was predicated on the assumption that apoplastic cross-linking of arabinoxylans in maize cell cultures is due to the action of endogenous peroxidases, but little was known about the apoplastic peroxidase activities involved or whether they were soluble or wall-bound. Here, we present further evidence in favour of a role for peroxidase in inter-arabinoxylan cross-linking and some characteristics of the specific peroxidase activities responsible.
|
RESULTS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
Hydrogen Peroxide Is Necessary But Not Sufficient for Cross-Linking of [3H]Arabinoxylans In Vivo
To establish an experimental system for studying the cross-linking of arabinoxylans in vivo, we incubated [pentosyl-3H]arabinoxylans (Mr
106) with maize cell-cultures for 24-h periods, starting 6–12 d after sub-culture. During these 24-h periods, <6% of the radioactivity (mean = 4%) became sequestered on or in the cells, indicating that only a small proportion of the polysaccharide substrate was removed from the (soluble, extracellular) fraction under observation—in contrast to the situation with the feruloyl-oligosaccharide studied by Encina and Fry (2005). In 6 and 7-day-old cultures, the [3H]arabinoxylans did not perceptibly change during a 24-h incubation; however, in 8-day-old cultures, they underwent a dramatic increase in Mr, indicating cross-linking (Figure 1, 
). In progressively older cultures, the cross-linking per 24 h diminished, but not to zero. Such age-related cross-linking of polysaccharides could be due to the peroxidase/peroxide-dependent oxidative coupling of feruloyl side-chains present on the [3H]arabinoxylan molecules. Evidence that the in-vivo cross-linking was H2O2-dependent was provided by the effect of added 20 mM KI, a peroxide scavenger, which strongly inhibited the cross-linking (Figure 1, •) and has been reported to suppress the bonding of A5f3X into the cell wall (Encina and Fry, 2005). KI at 20 mM is apparently not toxic to plant cells, since it can activate transcription (Anterola et al., 2002).
|
The peroxidase/peroxide-dependent oxidative coupling hypothesis predicts that the lack of cross-linking in 6 and 7-day-old control cultures (Figure 1,
) could be due to (1) a lack of H2O2, (2) a lack of peroxidase, or (3) the presence of antioxidants or other inhibitors of peroxidase action. To distinguish between these predictions, we supplied exogenous H2O2 to the cultures: this had no effect (Figure 1, 
), indicating that the absence of endogenous H2O2 was not the (sole) reason for the lack of cross-linking in 6 and 7-day-old cultures. The dose of H2O2 supplied (6 µM final concentration) is sufficient to cause extensive cross-linking in vitro (Burr and Fry, 2009). Thus, 6 and 7-day-old cultures must either lack suitable apoplastic peroxidases and/or possess apoplastic anti-oxidants such as the low-Mr inhibitors reported by Encina and Fry (2005). Peroxidase assays with o-dianisidine as electron donor demonstrated high peroxidase activity at all culture ages (data not shown), so the inhibitor hypothesis appears more likely.
Horseradish Peroxidase Blocks [3H]Arabinoxylan Cross-Linking In Vitro
In cell-free, dialysed culture filtrate from 8-d cultures, containing endogenous peroxidase (detectable by the o-dianisidine assay; see later), added H2O2 was sufficient to cause the cross-linking of [3H]arabinoxylans in vitro. Cross-linking was detectable within 10 min of the addition of H2O2, and continued for up to 24 h (Figure 2A). Surprisingly, however, type-II horseradish peroxidase (HRP) strongly inhibited H2O2-induced cross-linking in this system, while having relatively little effect in the absence of H2O2 (Figure 2B). Type-II HRP as dilute as 5 µg ml–1 had an appreciable inhibitory effect (Figure 2C). Type-VI HRP had a similar effect (data not shown).
|
In a comparison of type-II HRP with the endogenous maize peroxidase activity, we noted that the former at 5 µg ml–1 gave a rate of coloured product formation from o-dianisidine of approximately 0.7 A420 s–1, whereas 8-d maize culture filtrate gave a rate of 0.108 ± 0.007 A420 s–1. Thus, there was somewhat more HRP activity than maize peroxidase activity.
To test whether any esterase activity, possibly present as a contaminant in the HRP preparations, might have been removing the feruloyl groups from the arabinoxylans and thereby preventing polysaccharide cross-linking, we repeated the experiment with HRP but using [feruloyl-14C]arabinoxylans instead of [pentosyl-3H]arabinoxylans as substrate. Any free [14C]ferulic acid released by an esterase would exhibit Kav > 1.0; however, all the 14C remained associated with polysaccharides (Kav = 0.0–0.5; Figure 2D), indicating the absence of feruloylesterase activity. [Note: Kav is the elution volume relative to those of high-Mr dextran (Kav = 0) and glucose (Kav = 1).] In this experiment, the HRP again inhibited H2O2-induced cross-linking of arabinoxylans—though only partially, possibly because the [14C]arabinoxylans had a lower specific radioactivity than the [3H]arabinoxylans and therefore had to be used at a higher concentration (120 µM compared with 25 µM feruloyl residues in the reaction-mixture).
The inhibitory effect of HRP appeared to be unrelated to its enzymic activity, since HRP still largely blocked cross-linking (Figure 3B) after complete heat denaturation (as judged by the o-dianisidine assay; Figure 3A).
|
Direct evidence confirmed that maize cells secrete soluble peroxidase(s) that can cross-link arabinoxylans far more effectively than can HRP. We boiled a crude preparation of [3H]arabinoxylans for 15 min, partially denaturing the endogenous peroxidases, then treated it with HRP or with non-denatured (non-radioactive) culture filtrate, and/or H2O2 (Figure 4). Neither of the enzyme sources (HRP or culture filtrate) alone caused cross-linking of [3H]arabinoxylans. Exogenous H2O2 alone caused slight cross-linking, showing that 15 min at 100°C had not completely denatured the endogenous peroxidases. This residual H2O2-induced cross-linking was strongly enhanced by addition of non-denatured culture filtrate containing active peroxidases, supporting the view that the maize enzyme activity responsible for cross-linking was peroxidase. However, H2O2-induced cross-linking was abolished by HRP.
|
Culture filtrate from 8–12-day-old cultures consistently restored cross-linking activity to a boiled [3H]arabinoxylans preparation, and did not inhibit the cross-linking activity of an unboiled [3H]arabinoxylans preparation (data not shown), thus differing from HRP in both these characteristics.
Maize Cultures Secrete Relatively Heat-Stable and Heat-Labile Peroxidases—Only the Latter Cross-Link Arabinoxylans
The o-dianisidine assay readily detected peroxidase activity in culture filtrate of 8-day-old maize cell cultures and revealed that a proportion of this activity was remarkably resistant to heat-denaturation (Figure 5A). The kinetics of denaturation suggest that multiple peroxidases were present in the medium: the majority (95%) was completely denatured by 10 min boiling (like HRP; Figure 3A), whereas a minority (5%) was heat-stable for at least 30 min. Comparing 10 min boiling with 30 min boiling (Figure 5), we note that the o-dianisidine peroxidase activity decreased by only 35% during this additional 20-min period of boiling (Figure 5A) whereas the polysaccharide cross-linking activity disappeared completely (Figure 5B). Thus, the heat-stable peroxidase activity, detectable with o-dianisidine as substrate, is not capable of cross-linking arabinoxylans.
|
Role of Ionic Strength in Control of Cross-Linking
In-vitro cross-linking experiments using dialysed culture filtrate as the source of peroxidase were routinely conducted in 100 mM acetate (Na+) buffer, pH 4.7 (Figures 2–5). Varying the pH of the buffer to 3.7 (100 mM formate, Na+) or 7.2 (100 mM phosphate, Na+) had a small effect on cross-linking, the lowest pH being most effective; however, omission of buffer completely prevented cross-linking (Figure 6A). The measured pH of an unbuffered reaction mixture was
5; therefore, the buffer requirement did not appear to be principally related to the control of pH. The concentration of buffer necessary for cross-linking was very low: between 1 and 100 mM acetate, the endogenous peroxidase was fully active in catalyzing cross-linking, but activity was partially lost at lower concentrations (Figure 6B). This was an effect of ionic strength rather than a pH effect, as shown by the observation that cross-linking still occurred if the buffer salt was replaced by 1–100 mM NaCl, which has no buffering capacity (data not shown). We conclude that the oxidative coupling of arabinoxylans is not strongly pH-dependent within the physiological range, but requires the medium to have a minimum ionic strength.
|
|
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
It is now well established that living plant cells can oxidatively couple endogenous arabinoxylan-bound feruloyl groups in three cellular compartments: intraprotoplasmic (probably in Golgi cisternae), wall-bound, and soluble extracellular (Fry et al., 2000; Mastrangelo et al., 2009). In addition, exogenous feruloyl-oligosaccharides (e.g. A5f3X) infiltrated into the walls of living plant cells are oxidatively coupled to form chemically similar products (Encina and Fry, 2005). The products are diferulate groups and (predominantly) larger oligomers, which, at least in the case of arabinoxylans, bring about polysaccharide–polysaccharide cross-linking (Burr and Fry, 2009). Chemically similar oxidative coupling of feruloyl-arabinoxylans can be achieved in vitro by use of exogenous peroxidases + H2O2 (Geissmann and Neukom, 1971; Figueroa-Espinoza et al., 1999; Schooneveld-Bergmans et al., 1999), but it has been more difficult to prove that the endogenous agents responsible for the process in vivo are also peroxidases + H2O2. Alternative possibilities include laccase + O2, or non-enzymic reactions involving any of various reactive oxygen species (ROS).
The present paper supports the idea that apoplastic feruloyl coupling is dependent on endogenous H2O2 (scavenged by KI; Figure 1) and on endogenous enzymes (denatured by boiling; Figure 5B). Enzymes that use H2O2 to effect an oxidation are defined as peroxidases (E.C. 1.11.1.–). The occurrence, though not strictly action (Fry, 2004b), of apoplastic peroxidases is well established (e.g. Liu and Lamport, 1974; Price et al., 2003). We now conclude that the natural apoplastic coupling of arabinoxylans is principally peroxidase/H2O2-driven.
There is a small amount of cross-linking that is not inhibited by KI, peaking in 10-day-old cultures (Figure 1). This probably does not indicate a role for oxidase action in arabinoxylan cross-linking in older cultures, since dialysed culture filtrates from such cultures supported no cross-linking of arabinoxylans until H2O2 was added. If KI is indeed an effective scavenger of H2O2, a possible alternative explanation is a limited role for cell-generated ROS, such as superoxide or hydroxyl radicals, in an enzymic-independent cross-linking reaction.
Intraprotoplasmic coupling of polysaccharide-esterified feruloyl groups can also be monitored by use of in-vivo radiolabelling (Fry et al., 2000). However, intraprotoplasmic coupling cannot be tested for H2O2- or enzyme-dependence in the same way as in the present paper because the substrates and inhibitors cannot be reliably made to penetrate the plasma membrane. It thus remains to be tested whether the ferulate coupling that occurs prior to polysaccharide secretion is peroxidase/H2O2-dependent, as in the case of apoplastic coupling.
It is of interest to consider the factors potentially controlling the cross-linking of feruloyl-polysaccharides. If the cross-linking of arabinoxylans in the apoplast is predominantly peroxidase/H2O2-driven, it might be predicted that exogenous H2O2 would enhance the process. However, this was not the case (Figure 1): in particular, exogenous H2O2 did not initiate premature cross-linking. The observation that in 6 and 7-day-old cultures, no cross-linking occurred, despite the presence of peroxidase (endogenous soluble extracellular, readily detectable at all culture ages; data not shown), H2O2 and [3H]arabinoxylans (both exogenous), indicates the involvement of a competing sink for the H2O2. This sink could be the polar, heat-labile, low-Mr inhibitor of feruloyl-oligosaccharide cross-linking (Encina and Fry, 2005). This inhibitor was present in 8-day-old culture filtrate, which is usually prior to the appearance of cross-linked arabinoxylans (onset of cross-linking in Figure 1 was early; a more usual time-point is 9–11 d after sub-culture). The constitutive initiation of cross-linking might be related to the disappearance of the putative inhibitor, or the increased production of H2O2, or both.
The results of the in-vivo experiments were supported in vitro. In particular, a role for endogenous peroxidase in cross-linking was supported in the cell-free system. Native [3H]arabinoxylans in culture filtrate were dialysed, thus removing any endogenous ROS and the low-Mr inhibitor reported by Encina and Fry (2005), but leaving any endogenous soluble extracellular enzymes. In this solution, adding back H2O2 was sufficient to induce cross-linking, compatible with a peroxidase/H2O2-driven process. Also, as predicted, boiling destroyed the solution's ability to support H2O2-dependent cross-linking of arabinoxylans; thereafter, the ability to cross-link was restored by addition of (non-radioactive) non-denatured culture filtrate, which contained peroxidase activity.
Surprisingly, however, HRP did not restore the boiled culture-filtrate's ability to cross-link arabinoxylans; on the contrary, it further inhibited it (Figure 4). The apparent inhibition by HRP is surprising, since HRP/H2O2 has previously been reported to cause feruloylated polysaccharides to couple, even at lower peroxidase concentrations than used in the present study (20 µg ml–1 (Geissmann and Neukom, 1971) or 5 µg ml–1 (Schooneveld-Bergmans et al., 1999)). It seems unlikely that the HRP caused all the H2O2 to be diverted from reaction with arabinoxylans to reaction with some other substrate, since, in at least some of our experiments, the H2O2 was present at quite high concentrations (7.3 mM; Figure 2). It is very unlikely that any competing substrate could be present at >7.3 mM in dialysed culture filtrates, as would have been required to consume stoichiometrically all the available H2O2.
That the ‘inhibitory’ action of HRP was not enzymic was indicated by the finding that it still worked after 30 min boiling—which was sufficient to destroy all peroxidase activity as measured in the o-dianisidine assay (Figure 3). We suggest that the HRP preparation contains a substrate with which the arabinoxylans preferentially couple, precluding arabinoxylan–arabinoxylan cross-linking. Such an effect has been observed before on addition of low-Mr competing substrates such as sinapic or chlorogenic acid (Kerr and Fry, 2004).
One possibility is that the HRP itself constitutes such a substrate with which the arabinoxylans could couple, forming small HRP–arabinoxylan complexes (perhaps via tyrosine–ferulate bonds) instead of much larger arabinoxylan–arabinoxylan complexes (via ferulate–ferulate bonds). Since HRP has an Mr of only 4 x 104 (Welinder, 1979), the proposed arabinoxylan–HRP bonding would have no discernible effect on the size of the radiolabelled arabinoxylans (initial Mr 106) as determined by size-profiling on Sepharose CL-2B. HRP is certainly known to catalyze the oxidative coupling of free ferulic acid with the tyrosine residue of the tripeptide Gly–Tyr–Gly in a highly simplified model system (Oudgenoeg et al., 2002); and it is possible that it also catalyzes a similar reaction between feruloyl esters and polypeptides. The lowest effective HRP concentration used in our work (5 µg ml–1) corresponds to <2 µM Tyr residues, a concentration well below that of the feruloyl residues of the [3H]arabinoxylans (25 µM) or [14C]arabinoxylans (120 µM). Given this stoichiometry (<1:13 or <1:60), it would be necessary to propose that a very low ‘degree of substitution’ of arabinoxylans with coupled HRP prevents further arabinoxylan–arabinoxylan coupling. Previous workers studying HRP-catalyzed cross-linking of feruloyl polysaccharides used higher polysaccharide concentrations than we did (3–4 mg ml–1 (Schooneveld-Bergmans et al., 1999; Geissmann and Neukom, 1971); cf. <1 mg ml–1 in the present work), which may have been better able to out-compete the HRP. Our suggestion is thus that HRP can oxidatively couple itself to the feruloyl residues of arabinoxylans, whereas endogenous maize peroxidase(s) have evolved to prevent this problem, such as by having the tyrosine residues sterically hindered from reacting with the ferulates.
The endogenous peroxidase requires a minimum ionic strength in order to induce cross-linking (Figure 6B), and a small increase in cross-linking activity was noted with decreasing pH (Figure 6A). Both effects might be due to the presence of glucuronic acid residues in the arabinoxylan; deprotonated uronic acids would cause electrostatic repulsion between the chains, the pKa of glucuronic acid being 3.2. A role for ionic strength in the action of other cell wall-localized enzymes (e.g. xyloglucan endotransglucosylase activity) has been observed before (Takeda and Fry, 2004; Takeda et al., 2008). Ionic strength may be a widely important factor in controlling cell wall metabolism in vivo.
|
METHODS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
Cell Culture
Maize (Zea mays L. Black Mexican sweetcorn) cell-suspension cultures were grown under constant dim light on an orbital shaker at 25°C. Cells were sub-cultured fortnightly by removal of spent medium and addition of 30% of the cells into 200 ml fresh medium (0.47% (w/v) Murashige and Skoog basal inorganic medium (Sigma M5519), 2% (w/v) sucrose and 2 mg l–1 (w/v) 2,4-dichlorophenoxyacetic acid, pH 4.6–4.8), in 500-ml conical flasks. For radiolabelling experiments, 5-ml cultures were grown in loosely-capped 60-ml cylindrical beakers (3.7 cm internal diameter): under these conditions, the fresh weight typically increased exponentially from 0 to 5 d (
0.10–0.30 g culture–1), then linearly from 5 to 7 d (0.30–0.65 g culture–1); between 7 and 14 d, there was a slight decrease in fresh weight (0.65–0.55 g culture–1) (Kerr, 2002).
Production of Native Radiolabelled Soluble Extracellular Arabinoxylans
Maize cell-suspension cultures (5 ml) were incubated with [3H]arabinose (0.05 MBq, 148 GBq mmol–1) for 48 h starting 6 d after sub-culture. The cultures were then harvested, cells removed by filtration through nylon gauze, and low-Mr compounds removed from the spent medium by dialysis against water. Essentially all 3H in the dialysate was found in the pentose (arabinose and xylose) residues of non-cross-linked arabinoxylans (Burr and Fry, 2009). (Acid hydrolysis of similar, non-radioactive dialysates followed by paper chromatography revealed that the major components were arabinose and xylose, with smaller proportions of glucose and uronic acid.) The dialysed [pentosyl-3H]arabinoxylans were either used directly in incubation experiments or subjected to further treatments, such as boiling or alkali hydrolysis. Alkali hydrolysis of [pentosyl-3H]arabinoxylans was carried out in 1 M NaOH at room temperature for 48 h, followed by neutralization with 5 M acetic acid and dialysis against water.
[feruloyl-14C]Arabinoxylans were prepared as before (Burr and Fry, 2009).
[3H]Arabinoxylan Incubation with Maize Cell Cultures
In in-vivo cross-linking experiments, a solution containing 0.75 kBq of non-cross-linked [pentosyl-3H]arabinoxylans (Mr 106) was added to at least 20 vol. of a maize cell culture (6–12 d after sub-culture), and incubated for 24 h with or without additives (20 mM potassium iodide or 7.3 µM H2O2). The samples were then filtered and the culture filtrates subjected to gel-permeation chromatography.
Cell-Free System
Samples (90 µl) of dialysed 8-day-old culture filtrate containing native [pentosyl-3H]arabinoxylans (1.68 kBq ml–1) or [feruloyl-14C]arabinoxylans (0.2 kBq ml–1) were supplemented with 10 µl buffer (routinely acetate, Na+, pH 4.7, final concentration 100 mM), 10 µl H2O2 (final concentration 0–7300 µM), and 10 µl horseradish peroxidase (HRP) of type II (Sigma Chemical Co. catalogue no. P8250; lyophilized, salt-free powder) or type VI (P8375) (final concentration 0–80 µg ml–1) (final volume 120 µl). In some experiments, 30 µl of culture filtrate from maize cell cultures was also added (final volume 150 µl). Reactions were stopped by the addition of DTT to 10 mM.
Gel-Permeation Chromatography
Radiolabelled arabinoxylans were size-fractionated on a mini-column of Sepharose CL-2B column as before (Burr and Fry, 2009), with pyridine/acetic acid/0.5% aqueous chlorobutanol 1:1:98 (pH 4.7) as eluent, and Blue Dextran and glucose as markers. The void volume (V0) (Kav = 0.0) was indicated by the centre of the first (small) peak of Blue Dextran, and the included volume (Vi) (Kav = 1.0) by the glucose peak centre. Cross-linked [3H]arabinoxylans were eluted in the void volume.
Peroxidase Assay
Peroxidase solution (0.5 ml) was added to an assay mixture (1.0 ml) to give final concentrations of 850 µM o-dianisidine (3,3'-dimethoxybenzidine), 667 µM H2O2, 43 mM acetate (Na+, pH 4.7), and 6.7 mM NaH2PO4, and the increase of absorbance at 420 nm plotted. The activity of enzyme was determined from the slope in the initial linear part of the plot. For the boiling inactivation tests, HRP was diluted in de-ionized water to 1 µg ml–1, heated at 100°C for various times, then cooled and assayed with o-dianisidine.
|
FUNDING |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSFUNDING |
|---|
This work was supported by the Biotechnology and Biological Sciences Research Council (BB/C505791/1).
|
Acknowledgements |
|---|
We thank Mrs Janice Miller for excellent technical assistance. No conflict of interest declared.
|
Notes |
|---|
* The suggestion that diferulates can potentially form intra-polysaccharide loops has been debated. Hatfield and Ralph (1999) modelled in silico the oxidative coupling of an oligosaccharide of xylose [β-(1
4)-Xyl16] that carried two feruloyl–arabinose (Fer–Ara) groups variously spaced along this very short backbone. They concluded that, in most such structures, intra-chain looping was sterically implausible. However, the widest spacing tested was with the Fer–Ara groups on the nth and (n+5)th xylose residues—a very short span compared with real cell-wall arabinoxylans, which often have a backbone of several thousand xylose residues. Such a polymer backbone (chain length typically 2 µm) is 10x longer than the diameter of a Golgi vesicle, into which the polysaccharide had evidently fitted prior to secretion. This implies that in vivo, the polysaccharide chain must be capable of coiling, such that widely spaced Fer–Ara residues (located on, say, the nth and (n+2000)th xylose residues) could readily come into close contact, enabling diferulates to form intra-polysaccharide loops. We therefore do not accept Hatfield and Ralph's argument against intra-chain coupling. 
-
Allerdings E, Ralph J, Schatz PF, Gniechwitz D, Steinhart H, Bunzel M. Isolation and structural identification of diarabinosyl 8-O-4-dehydrodiferulate from maize bran insoluble fibre. Phytochemistry (2005) 66:113–124.[CrossRef][Web of Science][Medline]
Anterola AM, Jeon JH, Davin LB, Lewis NG. Transcriptional control of monolignol biosynthesis in Pinus taeda—factors affecting monolignol ratios and carbon allocation in phenylpropanoid metabolism. J. Biol. Chem. (2002) 277:18272–18280.
Brett CT, Wende G, Smith AC, Waldron KW. Biosynthesis of cell-wall ferulate and diferulates. J. Sci. Food Agric (1999) 79:421–424.[CrossRef][Web of Science]
Bunzel M, Funk C, Steinhart H. Semipreparative isolation of dehydrodiferulic and dehydrotriferulic acids as standard substances from maize bran. J. Sep. Sci. (2004) 27:1080–1086.[CrossRef][Web of Science][Medline]
Bunzel M, Ralph J, Brüning P, Steinhart H. Structural identification of dehydrotriferulic and dehydrotetraferulic acids isolated from insoluble maize bran fiber. J. Agric. Food Chem. (2006) 54:6408–6418.
Bunzel M, Ralph J, Funk C, Steinhart H. Structural elucidation of new ferulic acid-containing phenolic dimers and trimers isolated from maize bran. Tetrahedron Lett. (2005) 46:5845–5850.[CrossRef][Web of Science]
Burr SJ, Fry SC. Extracellular cross-linking of maize arabinoxylans by oxidation of feruloyl esters to form oligoferuloyl esters and ether-like bonds. Plant J (2009) 58:554–567.[CrossRef][Web of Science]
Encina A, Fry SC. Oxidative coupling of a feruloyl-arabinoxylan trisaccharide (FAXX) in the walls of living maize cells requires endogenous hydrogen peroxide and is controlled by a low-Mr apoplastic inhibitor. Planta (2005) 223:77–89.[CrossRef][Web of Science][Medline]
Fauré R, et al. An unambiguous nomenclature system for oligosaccharide motifs derived from heteroxylans found in plant cell walls. Aust. J. Chem. (2009) 62:533–537.[CrossRef]
Figueroa-Espinoza MC, Morel MH, Surget A, Rouau X. Oxidative cross-linking of wheat arabinoxylans by manganese peroxidase: comparison with laccase and horseradish peroxidase: effect of cysteine and tyrosine on gelation. J. Sci. Food Agric (1999) 79:460–463.[CrossRef][Web of Science]
Fry SC. Incorporation of [14C]cinnamate into hydrolase-resistant components of the primary cell wall. Phytochemistry (1984) 23:59–64.[CrossRef][Web of Science]
Fry SC. The Growing Plant Cell Wall: Chemical and Metabolic Analysis (2000) reprint edn (Caldwell, New Jersey: The Blackburn Press).
Fry SC. Oxidative coupling of tyrosine and ferulic acid residues: intra and extra-protoplasmic occurrence, predominance of trimers and larger products, and possible role in interpolymeric cross-linking. Phytochem. Rev. (2004a) 3:97–111.[CrossRef]
Fry SC. Tansley Review: Primary cell wall metabolism: tracking the careers of wall polymers in living plant cells. New Phytol (2004b) 161:641–675.[CrossRef][Web of Science]
Fry SC, Willis SC, Paterson AE. Intraprotoplasmic and wall-localised formation of arabinoxylan-bound diferulates and larger ferulate coupling-products in maize cell-suspension cultures. Planta (2000) 211:679–692.[CrossRef][Web of Science][Medline]
Funk C, Ralph J, Steinhart H, Bunzel M. Isolation and structural characterisation of 8-O-4/8-O4- and 8–8/8-O4-coupled dehydrotriferulic acids from maize bran. Phytochemistry (2005) 66:363–371.[CrossRef][Web of Science][Medline]
Geissmann T, Neukom H. Vernetzung von Phenolcarbonsäureestern von Polysacchariden durch oxidative phenolische Kupplung. Helv. Chim. Acta (1971) 54:1108–1112.[CrossRef][Web of Science]
Grabber JH, Hatfield RD, Ralph J, Zon' J, Amrhein N. Ferulate cross-linking in cell walls isolated from maize cell suspensions. Phytochemistry (1995) 40:1077–1082.[CrossRef][Web of Science]
Guillon F, Thibault JF. Oxidative cross-linking of chemically and enzymatically modified sugar-beet pectin. Carbohydr. Polym (1990) 12:353–374.[CrossRef][Web of Science]
Harris PJ, Hartley RD. Phenolic constituents of the cell walls of monocotyledons. Biochem. Syst. Ecol (1980) 8:153–160.[CrossRef][Web of Science]
Harris PJ, Kelderman MR, Kendon MF, McKenzie RJ. Monosaccharide compositions of unlignified cell walls of monocotyledons in relation to the occurrence of wall-bound ferulic acid. Biochem. Syst. Ecol (1997) 25:167–179.[CrossRef][Web of Science]
Hartley RD, Harris PJ. Phenolic constituents of the cell walls of dicotyledons. Biochem. Syst. Ecol (1981) 9:189–203.[CrossRef][Web of Science]
Hatfield RD, Ralph J, Grabber JH. Cell wall cross-linking by ferulates and diferulates in grasses. J. Sci. Food. Agric (1999) 79:403–407.[CrossRef][Web of Science]
Ishii T. Isolation and characterization of a diferuloyl arabinoxylan hexasaccharide from bamboo shoot cell-walls. Carbohydr. Res. (1991) 219:15–22.[CrossRef][Web of Science][Medline]
Kerr EM. The synthesis, wall-binding and breakdown of hemicelluloses in maize cell-suspension cultures (2002) PhD thesis, The University of Edinburgh.
Kerr EM, Fry SC. Pre-formed xyloglucans and xylans increase in molecular weight in three distinct compartments of a maize cell-suspension culture. Planta (2003) 217:327–339.[Web of Science][Medline]
Kerr EM, Fry SC. Extracellular cross-linking of xylan and xyloglucan in maize cell-suspension cultures: the role of oxidative coupling. Planta (2004) 219:73–83.[CrossRef][Web of Science][Medline]
Lindsay SE, Fry SC. Control of diferulate formation in dicotyledonous and gramineous cell-suspension cultures. Planta (2008) 227:439–452.[Web of Science][Medline]
Liu EH, Lamport DTA. An accounting of horseradish peroxidase isozymes associated with the cell wall and evidence that peroxidase does not contain hydroxyproline. Plant Physiol (1974) 54:870–876.
Mares DJ, Stone BA. Studies on wheat endosperm II. Properties of the wall components and studies on their organization in the wall. Aust. J. Biol. Sci. (1973) 26:813–830.
Markwalder HU, Neukom H. Diferulic acid as a possible cross-link in hemicelluloses from wheat germ. Phytochemistry (1976) 15:836–837.[CrossRef][Web of Science]
Mastrangelo L, Lenucci M, Piro G, Dalessandro G. Evidence for intra- and extra-protoplasmic feruloylation and cross-linking in wheat seedling roots. Planta (2009) 229:343–355.[CrossRef][Web of Science][Medline]
Micard V, Thibault J-F. Oxidative gelation of sugar-beet pectins: use of laccases and hydration properties of the cross-linked pectins. Carbohydr. Polym (1999) 39:265–273.[CrossRef][Web of Science]
Morrison IM. Extraction of hemicelluloses from plant cell-walls with water after preliminary treatment with methanolic sodium methoxide. Carbohydr. Res (1977) 57:C4–C6.
O'Neill MA, Ishii T, Albersheim P, Darvill AG. Rhamnogalacturonan II: Structure and function of a borate cross-linked cell wall pectic polysaccharide. Annu. Rev. Plant Biol. (2004) 55:109–139.[CrossRef][Medline]
Oudgenoeg G, Dirksen E, Ingemann S, Hilhorst R, Gruppen H, Boeriu CG, Piersma SR, van Berkel WJH, Laane C, Voragen AGJ. Horseradish peroxidase-catalysed oligomerization of ferulic acid on a template of a tyrosine-containing tripeptide. J. Biol. Chem. (2002) 277:21332–21340.
Pérez S, Rodríguez-Carvajal MA, Doco T. A complex plant cell wall polysaccharide: rhamnogalacturonan II: A structure in quest of a function. Biochimie (2003) 85:109–121.[CrossRef][Web of Science][Medline]
Price NJ, Pinheiro C, Soares CM, Ashford DA, Ricardo CP, Jackson PA. A biochemical and molecular characterisation of LEP1, an extensin peroxidase from Lupin. J. Biol. Chem. (2003) 278:41389–41399.
Proseus TE, Boyer JS. Calcium pectate chemistry causes growth to be stored in Chara corallina: a test of the pectate cycle. Plant Cell. Env (2008) 31:1147–1155.[Medline]
Ralph J, Quideau S, Grabber JH, Hatfield RD. Identification and synthesis of new ferulic acid dehydrodimers present in grass cell walls. J. Chem. Soc. Perkin Trans. (1994) 1:3485–3498.
Rombouts FM, Thibault J-F. Feruloylated pectic substances from sugar-beet pulp. Carbohydr. Res. (1986) 154:177–187.[CrossRef][Web of Science]
Rouau X, Cheynier V, Surget A, Gloux D, Barron C, Meudec E, Louis-Montero J, Criton M. A dehydrotrimer of ferulic acid from maize bran. Phytochemistry (2003) 63:899–903.[CrossRef][Web of Science][Medline]
Saulnier L, Crépeau MJ, Lehaye M, Thibault J-F, Garcia-Conesa MT, Kroon PA, Williamson G. Isolation and structural determination of two 5,5'-diferuloyl oligosaccharides indicate that maize heteroxylans are covalently cross-linked by oxidatively coupled ferulates. Carbohydr. Res. (1999) 320:82–92.[CrossRef][Web of Science]
Schooneveld-Bergmans MEF, Dignum MJW, Grabber JH, Beldman G, Voragen AGJ. Studies on the oxidative cross-linking of feruloylated arabinoxylans from wheat flour and wheat bran. Carbohydr. Polym (1999) 38:309–317.[CrossRef][Web of Science]
Takeda T, Fry SC. Control of xyloglucan endotransglucosylase activity by salts and anionic polymers. Planta (2004) 219:722–732.[Web of Science][Medline]
Takeda T, Miller JG, Fry SC. Anionic derivatives of xyloglucan function as acceptor but not donor substrates for xyloglucan endotransglucosylase activity. Planta (2008) 227:893–905.[CrossRef][Web of Science][Medline]
Waldron KW, Parr AJ, Ng A, Ralph J. Cell-wall esterified phenolic dimers: identification and quantification by reverse-phase high performance liquid chromatography and diode array detection. Phytochem. Analysis (1996) 7:305–312.
Ward G, Hadar Y, Bilkis I, Konstantinovsky L, Dosoretz CG. Initial steps of ferulic acid polymerization by lignin peroxidase. J. Biol. Chem. (2001) 276:18734–18741.
Welinder KG. Amino acid sequence studies of horseradish peroxidase. Eur. J. Biochem (1979) 96:483–502.[Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||