Loss of Highly Branched Arabinans and Debranching of Rhamnogalacturonan I Accompany Loss of Firm Texture and Cell Separation during Prolonged Storage of Apple1 Marı´a J. Pen˜a2 and Nicholas C. Carpita* Department of Botany and Plant Pathology, Purdue University, West Lafayette, Indiana 47907–2054
Growth and maturation of the edible cortical cells of apples (Malus domestica Borkh) are accompanied by a selective loss of pectin-associated (1/4)-b-D-galactan from the cell walls, whereas a selective loss of highly branched (1/5)-a-L-arabinans occurs after ripening and in advance of the loss of firm texture. The selective loss of highly branched arabinans occurs during the overripening of apples of four cultivars (Gala, Red Delicious, Firm Gold, and Gold Rush) that varied markedly in storage life, but, in all instances, the loss prestages the loss of firm texture, measured by both breaking strength and compression resistance. The unbranched (1/5)-linked arabinans remain associated with the major pectic polymer, rhamnogalacturonan I, and their content remains essentially unchanged during overripening. However, the degree of rhamnogalacturonan I branching at the rhamnosyl residues also decreases, but only after extensive loss of the highly branched arabinans. In contrast to the decrease in arabinan content, the loss of the rhamnogalacturonan I branching is tightly correlated with loss of firm texture in all cultivars, regardless of storage time. In vitro cell separation assays show that structural proteins, perhaps via their phenolic residues, and homogalacturonans also contribute to cell adhesion. Implications of these cell wall modifications in the mechanisms of apple cortex textural changes and cell separation are discussed.
Loss of firm texture and cell adhesion are ripeningassociated processes that affect fruit quality and postharvest storage. Because the pectin substances are enriched in the walls of fruit cells compared to primary walls of other growing cells and because they constitute major components of the middle lamellae, these acidic polysaccharides have long been suspected to play key roles in fruit ripening (Jarvis, 1984; Brummel and Harpster, 2001). Over the years changes in pectin composition during fruit ripening have been studied extensively in many species, but the association of biochemical changes with changes in fruit texture is still unclear. Pectin depolymerization has little impact on wall swelling and softening during tomato (Lycopersicon esculentum) fruit ripening (Smith et al., 1988), and in other instances the swelling state of several fruit walls may change markedly without significant pectin depolymerization (Redgwell et al., 1997). Rose et al. (1998) show that wall swelling and softening may be accompanied by pectin depolymerization events, but the events are either not the only requirements or they may be unrelated to ripening. One of the surprises that emerged from the completion of the Arabidopsis genome sequence is the large 1 This work was supported by a grant from the U.S.-Israel Bi-National Research and Development Fund (BARD) and by the Indiana 21st Century Research and Technology Fund. 2 Present address: Complex Carbohydrate Research Center, 315 Riverbend Road, University of Georgia, Athens, GA 30602–4712. * Corresponding author; e-mail
[email protected]; fax 765– 494–0363. Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.104.043679.
number of genes that putatively encode polysaccharide degrading enzymes (Arabidopsis Genome Initiative, 2000). For example, Arabidopsis contains more than 250 genes that encode enzymes that are related to those that depolymerize pectins (Henrissat et al., 2001). This significant allotment of the genome toward pectin hydrolysis forces plant biologists to reconsider subtle roles of pectins in wall swelling and softening. Most cell walls of growing tissues from edible fruit have a composition and architecture characteristic of the Type I primary cell wall, a wall generally found among all dicotyledonous and about one-half of the monocotyledonous species (Carpita and Gibeaut, 1993). Type I walls consist of xyloglucan-cellulose microfibril networks embedded in a matrix of several complex pectic polysaccharides. Two major pectic polysaccharide backbones are homogalacturonans (HGs) and rhamnogalacturonan Is (RG Is). Although the HGs are generally unbranched, xylosyl residues may be attached at the O-3 positions of the GalA residues to form xylogalacturonans (Schols et al., 1995). Four complex side chains may be added in a precise configuration to form RG IIs (O’Neill et al., 1996). RG IIs may dimerize through formation of boron di-diesters, which promote normal shoot development (Ishii et al., 2001; O’Neill et al., 2001) and contributes to the tensile strength of the wall (Ryden et al., 2003). Attached to the O-4 position of rhamnose residues of RG I are neutral sugar polymers of (1/5)a-L-arabinans with different degrees of branching at the O-3 and O-2 positions, (1/4)-b-D-galactans, and type I arabino-(1/4)-galactans (Brett and Waldron, 1996; Willats et al., 2001). These neutral side chains of the RG I in the cell wall are spatially regulated in
Plant Physiology, July 2004, Vol. 135, pp. 1305–1313, www.plantphysiol.org 2004 American Society of Plant Biologists
1305
Pen˜a and Carpita
relation with cell wall architecture and cell development in several plant tissues (Jones et al., 1997; Willats et al., 1999; Orfila and Knox, 2000; Bush et al., 2001). Developmental regulation of pectin composition may impact mechanical properties (Willats et al., 2001), and the dynamics of pectin fine structure may still constitute the missing link in defining the biochemical basis of fruit softening. Redgwell et al. (1997) found no correlation between the loss of Gal and the swelling and softening of several kinds of fruit. However, the Cnr tomato fruit ripening mutant, which has a disrupted deposition of (1/5)-a-L-arabinans, exhibits altered tissue properties (Orfila et al., 2001). The edible portion of apple (Malus domestica Borkh) is not a true fruit but fleshy tissue derived from floral tube (hypanthium; Esau, 1977). Similar to true fruit, the firm cortex of apple showed a marked loss of Gal and Ara from the side chains of the pectin polymers during the development and ripening. However, these events occur at different stages, with Gal loss during fruit development and Ara loss during ripening and overripening (Redgwell et al., 1997). The cortical tissues of ripe apples maintain wall firmness, but during late stages of ripening the tissues soften and cells separate. Here, we report a study of the chronology of the biochemical events associated with loss in firm texture and cell separation in four apple cultivars that vary markedly in their storage behavior. These studies show that the loss of highly branched 5-linked arabinans prestages the loss of texture but that loss of RG I branching after the loss of the arabinans coincides with the textural changes. Cell wall proteins, phenolic residues, and HGs are each implicated in cell separation that accompanies the texture loss.
RESULTS
Figure 1. Growth characteristics of Red Delicious apple fruit at different stages of development and ripening. A, Representative examples of apple fruit during the growth season and in the overripened state (27 weeks). Values are weeks after flowering. B, Weight and volume of apples harvested at each of these stages. Values are the mean 6 SD of 10 representative samples.
Growth of Red Delicious Apples
The pattern of growth of Red Delicious apples is typical of pomes (Fig. 1, A and B). These apples enlarge quickly during the first 3 to 4 weeks, with growth a result of both cell division and extension, but after 1 month, growth is primarily a result of cell expansion (Tukey and Young, 1942). During the second month apple growth continues by cell expansion, and at about 15 weeks the color of the epidermis changes from green to red and the flesh color shifts from green to white, both indicators of pending ripeness (Fig. 1A). Red Delicious apples ripen between 20 and 21 weeks. The flesh color and texture turns from white and firm to yellow, and the tissues soften during the transition to the overripe stage. Cell separation occurs about 6 weeks after ripening coincident with loss of tissue volume and integrity. Apple cultivars vary considerably in length of the growth and ripening stages. Those in this study varied from Gala, which ripens between 18 and 19 weeks after flowering, to Firm Gold, which ripens between 22 and 23 weeks, 1306
and Gold Rush, which ripens 26 to 27 weeks after flowering. Monosaccharide Composition and Linkage Structure of Apple Cell Walls during Development and Ripening
Ara (approximately 30%), Gal (approximately 20%), and GalA (approximately 20%) are the major monosaccharides in the walls of rapidly enlarging apples (cv Red Delicious). Gal and Ara levels decrease drastically during development but at distinctly different stages. Gal is lost primarily during the cell enlargement phase and is at its lowest levels upon maturation (approximately 8 mol%), whereas Ara is at its highest levels during maturation and ripening (approximately 25 mol%) and is lost primarily upon subsequent storage at ambient temperature. Linkage analysis shows that galactosyl residues are mostly nonreducing terminal and 4-linked, the latter indicating unbranched Plant Physiol. Vol. 135, 2004
Apple Cell Wall Dynamics during Storage
(1/4)-galactans (Fig. 2). The loss of Gal during the apple development is mostly from these pectic galactans. Arabinofuranosyl residues are primarily in 5-linked arabinans, but a substantial amount of branched arabinans also exist, by substitution of the 5-linked residues primarily at the O-3 positions but also at the O-2 positions. Changes in the proportion of arabinosyl branch-point residues parallel those of nonreducing terminal arabinosyl residues, whereas the amount of (1/5)-arabinan decreases only slightly (Fig. 2). Loss of Firm Texture during Storage
To determine the role of Ara in the softening that accompanies apple overripening, we studied four cultivars that varied markedly in storage life: Gala, which ripens earliest and exhibits rapid overripening, followed by later-ripening Red Delicious, Firm Gold, and Gold Rush, the latter of which ripens the latest in the season and exhibits a protracted shelf life. Texture analysis included determination of the breaking strength of tissue cylinders and resistance to compression of tissue plugs. These experiments allowed us to differentiate between physical behavior and physiological age in our correlations of the biochemical changes in polysaccharide structure. Texture analysis showed that these apples exhibit different firm textures at harvest and soften at different rates during the storage (Fig. 3). Of the four cultivars studied, Gala showed the least firm texture at harvest (Fig. 3). The force necessary to compress or break the apple tissue continuously decreased during the 6 weeks of ambient
temperature storage, after which cell separation became obvious. Red Delicious apples displayed firmer texture at harvest than Gala (Fig. 3). After the first week of storage, their breaking forces were lowered, whereas there was little change in the resistance to compression during the first 3 weeks of storage. Afterward, the firm texture was lost quickly, and loss of cell adhesion was observed after 6 weeks. A different behavior is found in Firm Gold apples (Fig. 3), which displayed firmer textures at harvest than Red Delicious and Gala but softened faster. In only 4 weeks the force required for breakage decreased to a level that was maintained during the remainder of storage. Gold Rush apples had the firmest texture at harvest and were capable of much longer storage (11 weeks without cell separation) than the other cultivars (Fig. 3). Gold Rush and Red Delicious apples maintained firm textures at 4C for more than 12 weeks. Gala apples softened slightly during the cold storage. On the other hand, Firm Gold apples conserved at 4C softened at the same rate as apples at ambient temperature but with a 1-week delay. Because the tissue compression assay is independent of the cell turgor and more dependent on cell separation, the similarity of compression and breaking strength tests indicated that cell turgor is not a principal factor in firm texture. After 5 week of storage, Gold Rush apples exhibited a noticeable loss of turgor, and the resultant increase in tissue flexibility made it impossible to break under the standardized conditions. However, loss of turgor was independent from the compression behavior of the tissue, and significant loss of texture was observed only after 6 weeks of storage. With the exception of Firm Gold, maintenance of firm texture was correlated with lateness of harvest. Changes in Composition of the Cell Wall during Apple Softening
Figure 2. Linkage distribution of selected arabinosyl and galactosyl residues in the cell walls of Red Delicious apples during the different stages of development and ripening. ¤, t-Ara; ), 3,5-Ara; :, 5-Ara; n, t-Gal; h, 4-Gal; d, 3,4-Gal. Bars represent variance of two samples. Variance did not extend beyond the symbol in values without bars. Absolute mol% of monosaccharides was determined by quantitative analysis of alditol acetate derivatives. Plant Physiol. Vol. 135, 2004
Given the marked differences in the storage behavior, we examined sugar composition and linkage structure to determine the principal changes that might be linked to the textural changes. Despite the differences of firm texture found between apples of the four cultivars, sugar analyses showed similar cell wall compositions at harvest. Ripe apple cell walls contained GalUA, Ara, and Xyl as predominant noncellulosic sugars (Table I). GalUa was the most abundant monosaccharide, with values between 32 and 40 total mol%. GalUA levels slightly decreased in the four cultivars during the storage, and for Red Delicious and Gold Rush they decreased more notably coincident with cell separation. Ara is the principal neutral sugar at this stage, with values between 24 and 27 mol%. The major event associated with the storage in the four cultivars was a marked decrease in Ara content that always preceded the loss of firm texture. The Ara contents also decreased before fruit softening in Firm Gold apples that softened early at 4C. On the other hand, Gal levels were very low and were maintained 1307
Pen˜a and Carpita
Gala, where the ratio of 4-GalA to 3,4-GalA remained constant. The major linked neutral sugar was 5-Ara, which accounted for about one-half of the total Ara in the cell wall. The remainder of the Ara was in the branch-point residues 3,5- and 2,5-Ara, with corresponding nonreducing t-Ara residues. The degree of branching decreased markedly during the apple softening, with concomitant loss of t-Ara residues. The decrease in total Ara is primarily a result of loss of these branched and terminal furanosyl residues (Fig. 4). Most of the Ara remaining in the cell wall upon loss of texture was 5-Ara. In all instances, the loss of branched Ara prestages the softening. Firm Gold apples soften faster than all others, and even soften during the cold storage, and even in this different behavior the loss of branched arabinans precedes fruit softening.
Figure 3. Changes in the force necessary to compress (left frames) or break (right frames) cylinders of cortex taken of Gala, Red Delicious, Firm Gold, and Gold Rush apples during the storage at 25C (s) and 4C (d).
without change through the storage period. Gold Rush apples had higher levels of Gal at harvest than the others, but the content decreased quickly during the first week of storage and then maintained levels observed in the other cultivars. Xyl content was between 14 and 19 mol% at harvest and increased relatively during storage. In the end of the storage period, Xyl became the predominant neutral sugar. Linkage Analysis
Linkage compositions of these apple cell walls were consistent with the predominant polysaccharides reported for ripe apples (Table I). Most of the GalA in ripe apples was 4-linked. The 3,4-GalA residues (4%– 6%) indicated the presence of a xylogalacturonans in these apples. Although total uronic acid content decreased slightly, 3,4-GalA increased relative to 4-GalA, indicating that xylogalacturonans enriched relative to HG during the apple softening. The exception was in 1308
Figure 4. Alterations in linkage structure and degree of RG I branching in cell walls from Gala, Red Delicious, Firm Gold, and Gold Rush apples during the storage at 25C (open symbols) and 4C (closed symbols). Bars represent variance of two samples. Variance did not extend beyond the symbol in values without bars. Absolute mol% of monosaccharides was determined by quantitative analysis of alditol acetate derivatives. Plant Physiol. Vol. 135, 2004
Apple Cell Wall Dynamics during Storage
Table I. Linkage analysis of cell walls from four cultivars at harvest and after storage to a similar stage of texture loss Linkage
Gala Harvest
Red Delicious 16 Weeks
Harvest
Firm Gold
16 Weeks
Gold Rush
Harvest
15 Weeks
Harvest
111 Weeks
2.7 0.2 2.4 2.2 4.5 0.4 tr 1.1 13.4 0.6 3.8 8.5 5.0 2.5 tr 1.9 0.4 4.9 2.2 0.3 1.1 0.6 tr tr tr tr 3.8 4.1 0.7 26.8 0.7 4.5
3.4 0.3 3.1 1.7 1.2 0.2 tr 0.6 11.8 0.4 0.7 11.1 5.1 2.1 tr 1.7 0.4 3.7 3.3 0.3 0.2 0.6 tr tr tr tr 4.0 4.4 1.1 28 0.8 9.1
2.2 tr 1.7 2.3 7.4 0.4 tr 0.5 9.3 0.3 6.3 8.3 3.5 2.0 0.1 1.8 0.4 5.8 2.8 0.3 3.9 0.8 tr 0.1 0.4 tr 3.7 4.2 0.4 24.1 0.8 5.7
2.9 tr 3.6 1.9 1.2 0.3 tr 0.6 10.3 0.2 0.9 13.0 5.6 3.5 tr 2.9 0.7 4.1 3.4 0.1 0.3 0.8 tr tr 0.3 tr 4.4 5.0 1.0 22.4 0.8 8.5
mol%
t-Fucp t-Rhap 2-Rhap 2,4-Rhap t-Araf t-Arap 2-Araf 3-Araf 5-Araf 2,5-Araf 3,5-Araf t-Xylp 2-Xylp 4-Xylp 2(3),4-Xylp 4-Manp 4,6-Manp t-Galp 2-Galp 3-Galp 4-Galp 6-Galp 3,4-Galp 3,6-Galp 4,6-Galp t-Glcp 4-Glcp 4,6-Glcp t-GalAp 4-GalAp 2,4-GalAp 3,4-GalAp
2.8 0.1 2.9 1.7 4.7 0.2 tr 0.9 14.1 0.6 4.3 10.1 5.8 3.3 tr 1.6 0.3 3.8 2.3 0.2 0.8 0.4 tr tr tr tr 2.8 3.2 0.6 27.5 0.6 4.1
4.2 tr 4.5 1.4 1.1 0.4 tr 0.7 13.5 0.4 0.5 14.6 4.8 3.1 tr 2.7 0.4 4.1 3.7 0.2 tr 0.5 tr tr tr tr 5.2 4.2 tr 24.2 0.4 4.5
2.0 tr 2.6 1.6 7.6 0.3 tr 0.7 11.6 1.0 6.0 7.0 4.1 3.2 tr 1.5 0.4 2.6 1.4 0.1 0.4 0.3 tr tr tr tr 2.6 2.9 tr 35.3 0.4 4.0
3.1 0.1 5.0 1.6 0.7 0.2 tr 0.4 9.2 tr 0.6 11.7 5.7 4.6 tr 2.9 0.7 3.7 2.7 0.1 0.1 0.3 tr tr tr 0.1 4.9 6.4 0.5 25.0 0.8 8.1
tr, Trace amounts less than 0.05%. For clarity, the variance is not shown but is less than 10% as observed in Figures 2 and 4. Absolute mol% of monosaccharides was determined by quantitative analysis of alditol acetate derivatives.
The degree of branching of the RG I was determined from the ratio of 2-Rha to 2,4-Rha (Fig. 4). At harvest, Gala and Red Delicious give ratios of about 0.6, and higher values were found for the firmer cultivars: 0.92 for Firm Gold and 1.38 for Gold Rush. The degree of branching decreased through the storage period in each of the four cultivars. This decrease was strongly correlated with the loss of apple firm texture in all cultivars. However, the decrease of degree of branching of the RG I was not accounted for by the decrease in branched arabinan side chains. About 50% of the total decrease in Gala and Red Delicious apples and between 25% to 30% in Firm Gold and Gold Rush occurred after the extensive loss of branched arabinans.
Determinants of Cell Adhesion and Separation
We tested different chemical solvents and enzymes that could mimic the cell separation in vitro to dePlant Physiol. Vol. 135, 2004
termine other cell wall components that could be implicated in the maintenance of cell adhesion. All four cultivars exhibited a similar behavior, and an example of this behavior in Gold Rush is shown in Figure 5. Separation was enhanced by CDTA, which disrupts the pectin association in the middle lamella. A similar response was obtained by oxidation of the phenolic substances with acidic sodium chlorite. Incubation with pectinase separated the tissue in clumps rather than effecting complete cell separation. Tissues maintain integrity after incubation for 2 h with water at ambient temperature, water at 70C, or proteinase K. The microcompression tests showed that CDTA, polygalacturonase, proteinase K, hot water, and acidic sodium chlorite each decreased tissue stiffness similarly in all four cultivars (Fig. 6A). Because sodium chlorite can oxidize phenolic residues in structural proteins, we determined if the proteinase K and the sodium chlorite had the same target molecule. Using Gold Rush as a model with the firmest texture suitable 1309
Pen˜a and Carpita
Figure 5. Cell separation degree in tissue from Gold Rush apples after treatment with different solutions: Water, CDTA, Proteinase K, Control NaClO2, NaClO2, and PGase. All scale bars 5 50 mm.
arabinans. In the ripe stage, Ara, Xyl, and GalA are the predominant noncellulosic sugars in cell walls of apples of the four cultivars. Similar results have been reported before for ripe apples (Aspinall and Fanous, 1984; Schols et al., 1995; Redgwell et al., 1997). Most of the GalA in ripe apples was 4-linked, but the 3,4-GalA residues and the corresponding nonreducing t-Xyl residues present indicated the presence of a xylogalacturonans in these apples, as has been reported (Schols et al., 1995). Xylogalacturonans enriched relative to HG during the apple overripening with the exception of Gala, where the relative proportions remain constant. This result suggests no association of the xylogalacturonans with the wall softening process. Curiously, the degrees of branching decrease in the overripe apples in all four cultivars, whereas the amounts of 5-linked arabinan remain essentially unchanged or decrease only slightly (Table I). These results indicate that two populations of arabinan, one with relatively unbranched (1/5)-a-L-Ara residues and the other highly branched, are present in different locations in the wall and exhibit different behaviors during ripening. The highly branched arabinan should be more susceptible to enzymatic degradation and extractions with the traditional methods for pectin
for sequential treatments, sodium chlorite treatment of tissues preincubated with proteinase K produced an additional softening of the tissue (Fig. 6B). On the other hand, Ca21 added to the incubation solution during the sodium chlorite treatment prevented part of the softening that occurred by heating the tissue but did not prevent the softening produced by treatment with sodium chlorite.
DISCUSSION
We found that the Gal- and Ara-rich side chains of pectins are extensively degraded in the cell wall of Red Delicious apple during development and ripening, but at distinctly different stages. Gal was mostly lost from the 4-linked galactan side chains of the RG I before fruit maturity, whereas loss of Ara is associated with the ripening, especially with the onset of the overripened state. The deposition of (1/4)-b-D-galactan is associated with the increase of firmness of pea cotyledons (McCartney et al., 2000). In intact Arabidopsis seedlings, the (1/4)-b-D-galactan epitope occurs in a restricted manner in epidermal, cortical, and endodermal cell walls and marks the transition zone at or near the onset of rapid cell elongation (McCartney et al., 2003). In apples, galactan is lost primarily during the cell enlargement phase, and the lowest levels were found upon maturation (Fig. 2). After this stage the neutral side chains attached to the RG I are basically 1310
Figure 6. Differences in the force necessary to compress the cell walls from the four cultivars of apple after chemical and enzymatic treatments. A, Treatment with (from left to right) water, proteinase K, CDTA, pectinase; hot water control NaClO2; NaClO2. B, After NaClO2 treatment compared to ambient and hot water controls (from left to right); standard incubation, previously treated with proteinase K, standard incubation plus CaCl2. Plant Physiol. Vol. 135, 2004
Apple Cell Wall Dynamics during Storage
isolation, and their structure and extractability are consistent with behavior of these pectic arabinans reported previously (Aspinall and Fanous, 1984; Redgwell et al., 1997; Perez et al., 2000). The linear (1/5)-a-Larabinans that are tightly associated with cellulose and less accessible for enzymes may be responsible for the high amount of Ara that remains in the alkali-insoluble a-cellulose residue in apple (Watt et al., 1999), even after endopolygalacturonase treatment of this residue (Redgwell et al., 1997). Arabinans are attached to the O-4 position of rhamnose in the RG I (Brett and Waldron, 1996). The extensive loss of short, highly branched arabinan chain may be responsible for the initial decrease observed of the degree of branching of RG I (Fig. 4). However, the degree of branching of RG I continually decreased through the storage period in each of the four cultivars, indicating that other side branches were present. Our results and those from studies of other fruit (Redgwell et al., 1997; Watt et al., 1999) demonstrate that other arabinans are tightly integrated in the wall. The Cnr tomato mutant shows a disrupted deposition of (1/5)-a-L-arabinan, a modified HG network with reduced calcium-binding capacity in the middle lamellae, and a reduced cell adhesion concomitant with a failure to swell (Orfila et al., 2001, 2002). A similar phenomenon is observed in the arabinan-rich walls of Commelina guard cells, in which the wall shape can be frozen open or closed by digestion of 5-linked arabinans (Jones et al., 2003). Cell walls from the Cnr mutant contain larger amounts of galactosyl- and arabinosylrich polysaccharides tightly bound in the cell wall. Unfortunately, there are no data on the linkage structure of the arabinan in the Cnr mutant to surmise if failure to remove the branched arabinan is responsible for the change of texture observed in this fruit. Texture analysis of apple during overripening allowed us to correlate between the physical behavior and physiological age and the biochemical changes in polysaccharide linkage structure. These comparisons reveal that the loss of the highly branched arabinan prestaged the loss of texture in all instances (Fig. 4). Since arabinan loss is almost complete before softening, it is unlikely to be directly responsible but perhaps a prerequisite to permit other wall components to become susceptible to enzymatic modification. Loss of firm texture in apple is strongly correlated with the debranching of RG I (e.g. Figs. 3 and 4). The debranching of RG I is unrelated to the earlier loss of the highly branched arabinans, so both of the major neutral side groups of RG I, arabinan, and galactan have been hydrolyzed at the time of the greatest loss of RG I branch-point residues. Vincken et al. (2003) suggest a model where HGs are attached to RG I at the O-4 position of rhamnose residues. One must caution that there is no precedent in the literature for the GalA(1/4)-Rha linkage, although HG attachment could be through Ara or Gal intermediates, or even through more labile ester linkages rather than glycosidic linkages (Kim and Carpita, 1992). Plant Physiol. Vol. 135, 2004
The decrease in the highly branched arabinans that prestage the loss of texture may permit RG I-associated HG to be susceptible to enzymatic hydrolysis to effect cell separation. In the tobacco (Nicotiana plumbaginifolia) mutant nolac-H14, RG Is with low levels of associated arabinans were not retained in the cells walls and middle lamella (Iwai et al., 2001). The altered HG network and reduced cell adhesion observed in the Cnr tomato mutant could be a consequence of disrupted arabinan deposition in this tomato cell wall (Orfila et al., 2001, 2002). A role for arabinans in anchoring pectin in the wall was suggested by these authors, but the decreased abundance and altered location of arabinan side chain when RG I was fragmented in muro by a rhamnogalacturonan lyase in transgenic potato plants indicated that cross-linkage between arabinans and other components of the wall is unlikely (Oomen et al., 2002). However, if the arabinans were preventing enzymatic hydrolysis of HG, as we propose here, their absence would produce loss of pectin network integrity without any direct attachment. Fruit softening and loss of cell adhesion may involve changes in other minor cell wall components. In vitro cell separation (Fig. 5) and tissue softening (Fig. 6) can be produced by substances or enzymes that modify pectins, but oxidation of phenols is equally effective. Sodium chlorite has been shown to be an effective means to oxidize phenolic substances without appreciable change in the carbohydrate structure of the wall (Carpita, 1984). Calcium prevents softening of the tissue upon heat treatment but does not reverse tissue softening produced by sodium chlorite. Calciumchelating agents not only can disrupt junction zones of HG but also destabilize boron cross-linking of the RG II dimers (Kobayashi et al., 1999). These results suggest that at least two different components are associated with the stiffness of the cell wall: (1) pectin, whose hydrolysis produces cell separation, and (2) protein and/or phenolic compounds, which serve as crosslinkers. Oxidation of phenols by sodium chlorite produced additional softening of tissues upon proteinase K treatment. We are unable to deduce if the difference is because other phenolic residues than those in structural proteins are oxidized by the sodium chlorite treatment or if this treatment disrupts the protein network more effectively than the proteinase K. The presence and role of phenolic compounds in apple cell walls are somewhat controversial. Phenolic compounds are associated with apple pectin (Stevens and Selvendran, 1984; Kravtchenko et al., 1993), but Renard et al. (2001) found that soluble phenols from the cytosol may associate to the wall spontaneously upon homogenization. Previous experiments in our laboratory showed that diluted alkali released phenolic compounds from apple cell walls, but also that the amount of those phenols increased when the isolation of the wall started with any organic solvent (data not shown). We cannot determine if those phenols are the aromatic amino acids of proteins or are components of 1311
Pen˜a and Carpita
pectin network, and further work will be necessary to resolve this issue. MATERIALS AND METHODS Plant Materials Apples (Malus domestica cv Borkh) were grown at Purdue University Department of Horticulture experimental orchard, West Lafayette, Indiana. Apples were picked at different stages of development and ripening from the tree. Ripe fruit were kept at ambient temperature in darkness during overripening stages. Upon harvest, weights and volumes of individual apples were measured, and the fruit were then peeled, the outer parenchyma cut into small portions, and immediately frozen with liquid nitrogen. Tissue samples were stored at 280C until all had been collected. Apples of four cultivars, Red Delicious, Gala, Firm Gold, and Gold Rush, were picked at ripe stages from trees grown at the same farm. For each cultivar, the apple collection was divided in two, one was kept at ambient temperature and the other at 4C, until cell separation had occurred in those at ambient temperature. Two apples from each cultivar and storage condition were harvested weekly and used for the texture analysis. Immediately after the fruit was peeled, the outer cortex was prepared and stored as described above for biochemical analyses performed subsequently.
Texture Analysis Every 1 to 2 weeks, two apples of each cultivar and storage condition were harvested for compression and breaking strength assays with a TA-XT2i texture analyzer (Stable Microsystems, distributed by Texture Technologies, Scarsdale, NY). For the fracture force measurements, 9- 3 40-mm cylinders of apple cortex were removed from the fruit and placed on a stainless steel support with 2 anvils set 3 cm apart. Force was applied perpendicularly to the apple cylinder with a blunt-edged anvil descending at a creep rate of 0.5 mm s21. The force to break the cylinder was digitally recorded. Compression upon a cylindrical plug of apple cortex 1.8-cm diameter 3 1-cm thick resting on a flat platform was performed on tissues from the same apple used for fracture force measurements. Force was applied to the piece using a round probe, 1 cm in diameter, descending at a rate of 0.5 mm s21. The force required to compress the apple tissue to a depth of 5 mm was recorded.
Determination of Factors Involved in Cell Separation Cortex cylinders (6-mm diameter 3 5-mm long) or slices (6-mm diameter 3 1-mm thick) of cultivar Gold Rush were incubated in 80% (v/v) ethanol at 70C for 30 min and washed twice with fresh 80% hot ethanol. After rehydration of the tissues, the cylinders were incubated by gentle rocking for up to 24 h at ambient temperature in water, 50 mM CDTA, pH 6.5, pectinase from Aspergillus niger (Sigma, St. Louis) in 100 mM sodium acetate buffer, pH 4.0, and proteinase K in 50 mM Tris[HCl] buffer, pH 7.5. Other cylinders were incubated with sodium chlorite to selectively oxidize phenolic compounds without altering the polysaccharide composition (Carpita, 1984). Typically, the material was incubated in 40 mL of 0.34 M NaClO2 in 65 mM HOAc or water (as control) at 70C for 2 h and then incubated by gentle shaking at ambient temperature for an additional 22 h. In some instances CaCl2 was added to the chlorite solution or water. Alternatively, the cylinders were incubated with proteinase K before sodium chlorite treatment. After incubation the tissues were washed several times with water. The slices were stained with 0.02% aqueous toluidine blue O for 5 min, then washed with water. Before examination by microscopy, the slices were mounted in water on a glass slide, and the glass coverslip was moved gently left and right two times. The tissue was observed in the light microscope, and the degree of cell separation produced by the different treatments was compared. Microcompression tests were done as described above but using a flat probe 7 mm descending at a creep rate of 0.1 mm s21. The force required to puncture the apple cylinder (6-mm diameter 3 5-mm long) to a depth of 3 mm was recorded.
Cell Wall Preparation Frozen portions of the cortex were homogenized in 1% SDS in 100 mM Tris[HCl], pH 7.2, by Polytron (Brinkmann Instruments, Westbury, NY). The
1312
cell walls were collected on nylon cloth (Nitex, 47 mm-mesh; Tetko, Depew, NY) to filter out the starch grains and washed sequentially with ethanol and water. The material was then homogenized in acetone and washed with acetone and water. Finally, the material was homogenized in 100 mM sodium chloride and washed with 100 mM sodium chloride and then water. The washes were repeated until no starch was detected upon staining with iodine/ iodide solution as observed by brightfield microscopy. Negligible uronic acid containing material was detected in the SDS-soluble extract and subsequent washes of the cell wall material.
Determination of the Cell Wall Sugar Composition The carboxyl groups of uronosyl in the cell walls were activated with 1-cyclohexyl-3-(-2-morpholinyl-4-ethyl) carbodiimide (methyl-p-toluene sulfonate) powder and reduced with NaBD4 to their respective 6,6-didueterio sugars (Kim and Carpita, 1992, as modified by Carpita and McCann, 1996). Uronosyl-reduced wall materials (1–2 mg) were hydrolyzed in 1 mL of 2 M trifluoroacetic acid (TFA) containing 1 mmol myoinositol (internal standard) at 120C for 90 min, and the supernatant was then evaporated at 40C in a stream of air. The sugars were reduced with NaBH4, and alditol acetates were prepared as described previously (Gibeaut and Carpita, 1991). Derivatives were separated by gas-liquid chromatography on a 0.25-mm 3 30-m column of SP-2330 (Supelco, Bellefonte, PA). After an initial loading hold at 80C, the temperature was rapidly adjusted to 170C at 25C min21, then programmed to 240C at 5C min21 with 10-min hold at the upper temperature. Helium flow was 1 mL min21 with a splitless injection. The electron-impact mass spectrometry was with a Hewlett-Packard (Palo Alto, CA) MSD at 70 eV and a source temperature of 250C. The proportion of 6,6-dideuteriogalactosyl was calculated using m/z 187/189, 217/219, and 289/291 after correction of 13C spillover of undeuterated fragments according to the equation described by Kim and Carpita (1992).
Linkage Analysis Uronosyl-reduced wall materials were suspended in DMSO in sonication bath and after 2 h were per-O-methylated with 2.5 M n-buthyllitium in hexane and methyl iodide according to Gibeaut and Carpita (1991). The permethylated polymers were recovered after addition of water to the mixture and partitioning into chloroform. The chloroform extracts were washed five times with water, and the chloroform was evaporated. The methylated polymers were hydrolyzed in 2 M TFA for 90 min at 120C. The TFA was evaporated at 30C in stream of nitrogen gas, and the sugars were reduced with NaBD4 and acetylated. The partial methylated alditol acetates were separated on the column with similar conditions as the alditol acetate. After an initial loading hold at 80C, the temperature was rapidly adjusted to 170C at 25C min21, then programmed to 210C at 2C min21 and then to 240C at 5C min21 with 6-min hold at the upper temperature. Gas-liquid chromatography-electronimpact mass spectrometry analysis was used to verify all derivative structures, as described by Carpita and Shea (1989).
ACKNOWLEDGMENTS We thank Nicholas Rozzi, Department of Food Science, Purdue University, for his assistance in teaching us the operation of the texture analyzer, and Dr. Peter Hirst, Department of Horticulture, Purdue University, for the selection of apple cultivars with varied ripening textures and storage characteristics. This is journal paper Number 17,232 of the Purdue University Agricultural Experiment Station. Received March 28, 2004; returned for revision April 8, 2004; accepted April 8, 2004.
LITERATURE CITED Arabidopsis Genome Initiative (2000) Analysis of the genome sequence of the flowering plant Arabidopsis thaliana. Nature 408: 796–815 Aspinall GO, Fanous HK (1984) Structural investigations on the nonstarchy polysaccharides of apples. Carbohyd Polym 4: 193–214 Brett CT, Waldron KW (1996) Physiology and Biochemistry of Plant Cell Walls, Ed 2. Chapman and Hall, London
Plant Physiol. Vol. 135, 2004
Apple Cell Wall Dynamics during Storage
Brummel DA, Harpster MH (2001) Cell wall metabolism in fruit softening and quality and its manipulation in transgenic plants. Plant Mol Biol 47: 311–340 Bush MS, Marry M, Huxham IM, Jarvis MC, McCann MC (2001) Developmental regulation of pectic epitopes during potato tuberisation. Planta 213: 869–880 Carpita NC (1984) Fractionation of hemicelluloses from maize cell walls with increasing concentrations of alkali. Phytochemistry 23: 1089–1093 Carpita NC, Gibeaut DM (1993) Structural models of primary-cell walls in flowering plants: consistency of molecular structure with the physiological properties of the cell wall during the growth. Plant J 3: 1–30 Carpita NC, McCann MC (1996) Some new methods to study plant polyuronic acids and their esters. In R Townsend, A Hotchkiss, eds, Progress in Glycobiology. Marcel Dekker, New York, pp 595–611 Carpita NC, Shea EM (1989) Linkage structure by gas chromatographymass spectrometry of partially-methylated alditol acetates. In CJ Biermann, GD McGinnis, eds, Analysis of Carbohydrates by GLC and MS. CRC Press, Boca Raton, FL, pp 155–216 Esau K (1977) Fruit. In Anatomy of Seed Plants. John Wiley & Sons, NY Gibeaut DM, Carpita NC (1991) Tracing the biosynthesis of the cell wall in intact cells and plants. Selective turnover and alteration of cytoplasmic and cell wall polysaccharides of proso millet cells in liquid culture and Zea mays seedlings. Plant Physiol 97: 551–561 Henrissat B, Coutinho PM, Davies GJ (2001) A census of carbohydrateactive enzymes in the genome of Arabidopsis thaliana. Plant Mol Biol 47: 55–72 Ishii T, Matsunaga T, Hayashi N (2001) Formation of rhamnogalacturonan II-borate dimer in pectin determines cell wall thickness of pumpkin tissue. Plant Physiol 126: 1698–1705 Iwai H, Ishii T, Satoh S (2001) Absence of arabinan in the side chains of pectin polysaccharides strongly associated with cell walls Nicotiana plumbaginifolia non-organogenic callus with loosely attached constituent cells. Planta 213: 907–915 Jarvis MC (1984) Structure and properties of pectin gels in plant cell walls. Plant Cell Environ 7: 153–164 Jones L, Milne JL, Ashford D, McQueen-Mason SJ (2003) Cell wall arabinan is essential for guard cell function. Proc Natl Acad Sci USA 100: 11783–11788 Jones L, Seymour GB, Knox JP (1997) Localization of pectic galactan in tomato cell walls using a monoclonal antibody specific to (1/4)-b-Dgalactan. Plant Physiol 113: 1405–1412 Kim J-B, Carpita NC (1992) Changes in esterification of the uronic acids groups of cell wall polysaccharides during elongation of maize coleoptiles. Plant Physiol 98: 646–653 Kobayashi M, Nakagawa H, Asaka T, Matoh T (1999) Borate-rhamnogalacturonan II bonding reinforced by Ca21 retains pectic polysaccharides in higher-plant cell walls. Plant Physiol 119: 199–203 Kravtchenko TP, Penci M, Voragen AGJ, Plinik W (1993) Enzymatic and chemical degradation of some industrial pectins. Carbohyd Polym 20: 195–205 McCartney L, Ormerod AP, Gidley MJ, Knox JP (2000) Temporal and spatial regulation of pectic (1/4)-b-D-galactan in cell walls of developing pea cotyledons: implications for mechanical properties. Plant J 22: 105–113 McCartney L, Steele-King CG, Jordan E, Knox JP (2003) Cell wall pectic (1/4)-b-D-galactan marks the acceleration of cell elongation in the Arabidopsis seedling root meristem. Plant J 33: 447–454 O’Neill MA, Eberhard S, Albersheim P, Darvill AG (2001) Requirement of borate cross-linking of cell wall rhamnogalacturonan II for Arabidopsis growth. Science 294: 846–849 O’Neill MA, Warrenfeltz D, Kates K, Pellerin P, Doco T, Darvill AG,
Plant Physiol. Vol. 135, 2004
Albersheim P (1996) Rhamnogalacturonan-II, a pectic polysaccharide in the walls of growing plant cell, forms a dimer that is covalently crosslinked by a borate ester. In vitro conditions for the formation and hydrolysis of the dimer. J Biol Chem 271: 22923–22930 Oomen RJFJ, Doeswijk-Voragen CHL, Bush MS, Vincken J-P, Borkhardt B, van den Broek LAM, Corsar J, Ulvskov P, Voragen AGJ, McCann MC, et al (2002) In muro fragmentation of the rhamnogalacturonan I backbone in potato (Solanum tuberosum L.) results in a reduction and altered location of the galactan and arabinan side-chains and abnormal periderm development. Plant J 30: 403–413 Orfila C, Huisman MMH, Willats WGT, van Alebeek GJWM, Schols HA, Seymour GB, Knox JP (2002) Altered cell wall disassembly during ripening of Cnr tomato fruit: implications for cell adhesion and fruit softening. Planta 215: 440–447 Orfila C, Knox JP (2000) Spatial regulation of pectin polysaccharides in relation to pit fields in cell walls of tomato fruit pericarp. Plant Physiol 122: 775–781 Orfila C, Seymour GB, Willats WGT, Huxham IM, Jarvis MC, Dover CJ, Thompson AJ, Knox JP (2001) Altered middle lamella homogalacturonan and disrupted deposition of (1/5)-a-L-arabinan in the pericarp of Cnr, a ripening mutant of tomato. Plant Physiol 126: 210–221 Perez S, Mazeau K, Herve´ du Penhoat C (2000) The three-dimensional structures of the pectin polysaccharides. Plant Physiol Biochem 38: 37–55 Redgwell J, Fisher M, Kendal E, MacRae EA (1997) Galactose loss and fruit ripening: high-molecular-weight arabinogalactans in the pectin polysaccharides of fruit cell walls. Planta 203: 174–181 Renard CMGC, Baron A, Guyot S, Drilleau J-F (2001) Interactions between apple cell walls and native apple polyphenols: quantification and some consequences. Int J Biol Macromol 29: 115–125 Rose JKC, Hadfield KA, Labavitch JM, Bennett AB (1998) Temporal sequence of cell wall disassembly in rapidly ripening melon fruit. Plant Physiol 117: 345–361 Ryden P, Sugimoto-Shirasu K, Smith AC, Findlay K, Reiter W-D, McCann MC (2003) Tensile properties of Arabidopsis cell walls depend on both a xyloglucan cross-linked microfibrillar network and rhamnogalacturonan II-borate complexes. Plant Physiol 132: 1033–1040 Schols HA, Bakx EJ, Schipper D, Voragen AGJ (1995) A xylogalacturonan subunit present in the modified hairy regions of apple pectin. Carbohydr Res 279: 265–279 Smith CJS, Watson CF, Ray I, Bird CR, Morris PC, Schuch W, Grierson D (1988) Antisense RNA inhibition of polygalacturonase gene-expression in transgenic tomatoes. Nature 334: 724–726 Stevens BJH, Selvendran RR (1984) Structural features of cell wall polymers of the apple. Carbohydr Res 135: 155–166 Tukey HB, Young JO (1942) Gross morphology and histology of developing fruit of the apple. Bot Gaz 104: 3–25 Vincken JP, Schols HA, Oomen RJFJ, McCann MC, Ulvskov P, Voragen AGJ, Visser RGF (2003) If homogalacturonan were a side chain of rhamnogalacturonan I. Implications for cell wall architecture. Plant Physiol 132: 1781–1789 Watt DK, Brasch DJ, Larsen DS, Melton LD (1999) Isolation, characterisation, and NMR study of xyloglucan from enzymatically depectinised and non-depectinised apple pomace. Carbohyd Polym 39: 165–180 Willats WGT, McCartney L, Mackie W, Knox JP (2001) Pectin: cell biology and prospects for functional analysis. Plant Mol Biol 47: 9–27 Willats WGT, Steele-King CG, Markus SE, Knox JP (1999) Side chains of pectic polysaccharides are regulated in relation to cell proliferation and cell differentiation. Plant J 20: 619–628
1313
