Extensive solubilization and depolymerization of cell wall polysaccharides during avocado (Perseaamericana) ripening involves concerted action of polygalacturonase and pectinmethylesterase

During the ripening of avocado ( Persea americana Mill.) fruit, water-soluble polyuronides increased dramatically, concomitant with marked downshifts in molecular mass. Treatment of cell walls from pre-ripe fruit with purified avocado polygalacturonase (PG, EC 3.2.1.15) promoted the release and molecular mass downshift of polyuronides. The polyuronides released by PG were similar in size distribution to water-soluble polyuronides from fruit at intermediate stages of ripening. Polyuronides released from pre-ripe fruit by PG, although of relatively high molecular mass, were not further degraded upon additional incubation with fresh enzyme. Similarly, water-soluble polyuronides prepared from fruit at intermediate stages of ripening were largely resistant to the action of purified PG in vitro. When polyuronides derived from fruit at intermediate stages of ripening were treated with weak alkali or pectinmethylesterase (PME, EC 3.1.1.11), extensive molecular mass downshifts occurred in response to incubation with PG. These results suggest that PG plays the central role in polyuronide degradation in ripening avocado fruit cell walls and that partial de-esterification is necessary for the increase in the susceptibility of polyuronides to PG. Differences in the patterns of polyuronide depolymerization in avocado fruit compared with the more thoroughly characterized tomato fruit are discussed.     

Polyuronides in Avocado (Persea americana) and Tomato (Lycopersicon esculentum) Fruits Exhibit Markedly Different Patterns of Molecular Weight Downshifts during Ripening

Avocado (Persea americana) fruit experience a rapid and extensive loss of firmness during ripening. In this study, we examined whether the chelator solubility and molecular weight of avocado polyuronides paralleled the accumulation of polygalacturonase (PG) activity and loss in fruit firmness. Polyuronides were derived from ethanolic precipitates of avocado mesocarp prepared using a procedure to rapidly inactivate endogenous enzymes. During ripening, chelator (cyclohexane-trans-1,2-diamine tetraacetic acid [CDTA])-soluble polyuronides increased from approximately 30 to 40 [mu]g of galacturonic acid equivalents (mg alcohol-insoluble solids)-1 in preripe fruit to 150 to 170 [mu]g mg-1 in postclimacteric fruit. In preripe fruit, chelator-extractable polyuronides were of high molecular weight and were partially excluded from Sepharose CL- 2B-300 gel filtration media. Avocado polyuronides exhibited marked downshifts in molecular weight during ripening. At the postclimacteric stage, nearly all chelator-extractable polyuronides, which constituted from 75 to 90% of total cell wall uronic acid content, eluted near the total volume of the filtration media. Rechromatography of low molecular weight polyuronides on Bio-Gel P-4 disclosed that oligomeric uronic acids are produced in vivo during avocado ripening. The gel filtration behavior and pattern of depolymerization of avocado polyuronides were not influenced by the polyuronide extraction protocol (imidazole versus CDTA) or by chromatographic conditions designed to minimize interpolymeric aggregation. Polyuronides from ripening tomato (Lycopersicon esculentum) fruit extracted and chromatographed under conditions identical with those used for avocado polyuronides exhibited markedly less rapid and less extensive downshifts in molecular weight during the transition from mature-green to fully ripe. Even during a 9-d period beyond the fully ripe stage, tomato fruit polyuronides exhibited limited additional depolymerization and did not include oligomeric species. A comparison of the data for the avocado and tomato fruit indicates that downshifts in polyuronide molecular weight are a prominent feature of avocado ripening and may also explain why molecular down-regulation of PG (EC 3.2.1.15) in tomato fruit has resulted in minimal effects on fruit performance until the terminal stages of ripening.     

Methyl de-esterification as a major factor regulating the extent of pectin depolymerization during fruit ripening: a comparison of the action of avocado (Persea americana) and tomato (Lycopersicon esculentum) polygalacturonases

Pectinmethylesterase (PME, EC 3.2.1.11) and polygalacturonase (PG, EC 3.2.1.15) are known to operate in tandem to degrade methylesterified polyuronides. In this study, PGs purified from tomato and avocado fruit were compared in terms of their capacity to hydrolyze water-soluble polyuronides from avocado before and following enzymic or chemical de-esterification. When assayed using polygalacturonic acid or polyuronides from avocado fruit, the activity of PG from tomato fruit was 3-4 times higher than that from avocado fruit. High molecular mass, low methylesterified (33%) water-soluble polyuronides (WSP) from pre-ripe avocado fruit (day 0) were partially depolymerized upon incubation with purified avocado and tomato PGs. In contrast, middle molecular mass, highly methylesterified (74%) WSP from day 2 fruit were largely resistant to the action of both PGs. PME or weak alkali treatment of highly methylesterified WSP decreased the methylesterification values to 11 and 4.5%, respectively. Treatment of de-esterified WSP with either avocado or tomato PGs caused extensive molecular mass downshifts, paralleling those observed during avocado fruit ripening. Although PME and PG are found in many fruits, the pattern of depolymerization of native polyuronides indicates that the degree of cooperativity between these enzymes in vivo differs dramatically among fruits. The contribution of PME to patterns of polyuronide depolymerization observed during ripening compared with physically compromised fruit tissues is discussed.     

Effect of Antisense Suppression of Endopolygalacturonase Activity on Polyuronide Molecular Weight in Ripening Tomato Fruit and in Fruit Homogenates

Fruit of tomato (Lycopersicon esculentum Mill.) in which endopolygalacturonase (PG) activity had been suppressed to <1% of wild-type levels were slightly firmer than nontransgenic controls later in ripening. Enzymically inactive cell walls were prepared from these ripening fruit using Tris-buffered phenol. When extracted with chelator followed by Na2CO3, the amounts of pectin solubilized from cell walls of nontransgenic control or from transgenic antisense PG fruit were similar. Size-exclusion chromatography analysis showed that, relative to controls, in antisense PG fruit polyuronide depolymerization was delayed in the chelator-soluble fraction throughout ripening and reduced in the Na2CO3-soluble fraction at the overripe stage. Reduced pectin depolymerization rather than altered extractability thus may have contributed to enhanced fruit firmness. Substantially larger effects of suppressed PG activity were detected in tomato fruit homogenates processed to paste. In control paste the majority of the polyuronide was readily soluble in water and was very highly depolymerized. In antisense PG paste the proportion of polyuronide solubilized by water was reduced, and polyuronides retained a high degree of polymerization. The suppression of fruit PG activity thus has a small effect on polyuronide depolymerization in the fruit but a much larger effect in paste derived from these fruit. This indicates that in the cell wall PG-mediated degradation of polyuronide is normally restricted but that in tissue homogenates or in isolated cell walls this restriction is removed and extensive pectin disassembly results unless PG is inactivated.     

Role of membranes of mycelial Mucor rouxii in synthesis and secretion of cell wall matrix polymers.

Membrane fractions of the mycelial form of Mucor rouxii contained endogenous polyuronides and glycoproteins with sugar compositions similar to the cell wall polyuronide (mucoran) and extracellular (wall and filtrate) glycoproteins, respectively. The polyuronide pool was rapidly labeled with D-[U-14C]glucose and on pulse-chase experiments appeared to turn over, suggesting that these polymers were precursors of the wall mucoran. In contrast, the glycoproteins appeared to accumulate. Although the membrane-associated glycoproteins had molecular weights similar to those of the extracellular glycoproteins, the bulk of the membrane-associated polyuronides were of a molecular weight lower than that of the wall polyuronide.     

Degradation of Cell Wall Polysaccharides during Tomato Fruit Ripening

Changes in neutral sugar, uronic acid, and protein content of tomato (Lycopersicon esculentum Mill) cell walls during ripening were characterized. The only components to decline in amount were galactose, arabinose, and galacturonic acid. Isolated cell walls of ripening fruit contained a water-soluble polyuronide, possibly a product of in vivo polygalacturonase action. This polyuronide and the one obtained by incubating walls from mature green fruit with tomato polygalacturonase contained relatively much less neutral sugar than did intact cell walls. The ripening-related decline in galactose and arabinose content appeared to be separate from polyuronide solubilization. In the rin mutant, the postharvest loss of these neutral sugars occurred in the absence of polygalacturonase and polyuronide solubilization. The enzyme(s) responsible for the removal of galactose and arabinose was not identified; a tomato cell wall polysaccharide containing galactose and arabinose (6:1) was not hydrolyzed by tomato β-galactosidase.     

Cell Wall Metabolism in Ripening Fruit: IV. Characterization of the Pectic Polysaccharides Solubilized during Softening of ;Bartlett' Pear Fruit

Fractionation of pectic polysaccharides from the juice of ripening `Bartlett' pears (Pyrus communis) gave two general types of polyuronides. The major type was a homogalacturonan (HGA) whose molecular weight decreased upon ripening. The other type comprised heteropolymers composed of various amounts of arabinose, rhamnose, and galactose. Treatment of the major arabinose-containing heteropolymeric fraction of high molecular weight (400,000) with a pear exo-polygalacturonase to degrade contaminating HGA gave a polyuronide which was inert to tomato endopolygalacturonase. Glycosyl-linkage analysis of this arabinosyl-polyuronide gave results expected from a rhamnogalacturonan I-like polysaccharide with large, highly branched araban side chains (RG-I). A linkage between HGA and RG-I was not found. RG-I, in ripening pears, appeared to be degraded with the initial loss of much of its arabinose.     

Polygalacturonase Isozymes and Pectin Depolymerization in Transgenic rin Tomato Fruit

We have previously described the construction and expression of a chimeric gene that allows developmentally regulated expression of tomato (Lycopersicon esculentum) polygalacturonase in ripening-impaired, mutant (rin) tomato fruit (JJ Giovannoni, D DellaPenna, AB Bennett, RL Fischer [1989] The Plant Cell 1: 53-63). We now show that expression of the chimeric polygalacturonase gene in rin tomato fruit resulted in the accumulation of all three polygalacturonase isozymes (PG1, PG2A, and PG2B). Polyuronide solubilization and polyuronide depolymerization both reached their maximal levels in transgenic rin fruit prior to the appearance of PG2 isozymes. These results demonstrate that PG1, PG2A, and PG2B all arise by differential processing of a single gene product and further suggest that the PG1 isozyme is sufficient to carry out both polyuronide solubilization and depolymerization in vivo.     

Molecular size and separability features of pea cell wall polysaccharides : implications for models of primary wall structure

Relative molecular size distributions of pectic and hemicellulosic polysaccharides of pea (Pisum sativum cv Alaska) third internode primary walls were determined by gel filtration chromatography. Pectic polyuronides have a peak molecular mass of about 1100 kilodaltons, relative to dextran standards. This peak may be partly an aggregate of smaller molecular units, because demonstrable aggregation occurred when samples were concentrated by evaporation. About 86% of the neutral sugars (mostly arabinose and galactose) in the pectin cofractionate with polyuronide in gel filtration chromatography and diethylaminoethyl-cellulose chromatography and appear to be attached covalently to polyuronide chains, probably as constituents of rhamnogalacturonans. However, at least 60% of the wall's arabinan/galactan is not linked covalently to the bulk of its rhamnogalacturonan, either glycosidically or by ester links, but occurs in the hemicellulose fraction, accompanied by negligible uronic acid, and has a peak molecular mass of about 1000 kilodaltons. Xyloglucan, the other principal hemicellulosic polymer, has a peak molecular mass of about 30 kilodaltons (with a secondary, usually minor, peak of approximately 300 kilodaltons) and is mostly not linked glycosidically either to pectic polyuronides or to arabinogalactan. The relatively narrow molecular mass distributions of these polymers suggest mechanisms of co- or postsynthetic control of hemicellulose chain length by the cell. Although the macromolecular features of the mentioned polymers individually agree generally with those shown in the widely disseminated sycamore cell primary wall model, the matrix polymers seem to be associated mostly noncovalently rather than in the covalently interlinked meshwork postulated by that model. Xyloglucan and arabinan/galactan may form tightly and more loosely bound layers, respectively, around the cellulose microfibrils, the outer layer interacting with pectic rhamnogalacturonans that occupy interstices between the hemicellulose-coated microfibrils.     

Cell Wall Metabolism in Ripening Fruit (VI. Effect of the Antisense Polygalacturonase Gene on Cell Wall Changes Accompanying Ripening in Transgenic Tomatoes)

Cell walls of tomato (Lycopersicon esculentum Mill.) fruit, prepared so as to minimize residual hydrolytic activity and autolysis, exhibit increasing solubilization of pectins as ripening proceeds, and this process is not evident in fruit from transgenic plants with the antisense gene for polygalacturonase (PG). A comparison of activities of a number of possible cell wall hydrolases indicated that antisense fruit differ from control fruit specifically in their low PG activity. The composition of cell wall fractions of mature green fruit from transgenic and control (wild-type) plants were indistinguishable except for trans-1,2-diaminocyclohexane-N,N,N[prime],N[prime]-tetraacetic acid (CDTA)-soluble pectins of transgenic fruit, which had elevated levels of arabinose and galactose. Neutral polysaccharides and polyuronides increased in the water-soluble fraction of wild-type fruit during ripening, and this was matched by a decline in Na2CO3-soluble pectins, equal in magnitude and timing. This, together with compositional analysis showing increasing galactose, arabinose, and rhamnose in the water-soluble fraction, mirrored by a decline of these same residues in the Na2CO3-soluble pectins, suggests that the polyuronides and neutral polysaccharides solubilized by PG come from the Na2CO3-soluble fraction of the tomato cell wall. In addition to the loss of galactose from the cell wall as a result of PG activity, both antisense and control fruit exhibit an independent decline in galactose in both the CDTA-soluble and Na2CO3-soluble fractions, which may play a role in fruit softening.     

Multiple forms of ADP-glucose pyrophosphorylase from tomato fruit.

ADP-glucose pyrophosphorylase (AGP) was purified from tomato (Lycopersicon esculentum Mill.) fruit to apparent homogeneity. By sodium dodecyl sulfate-polyacrylamide gel electrophoresis the enzyme migrated as two close bands with molecular weights of 50,000 and 51,000. Two-dimensional polyacrylamide gel electrophoresis analysis of the purified enzyme, however, revealed at least five major protein spots that could be distinguished by their slight differences in net charge and molecular weight. Whereas all of the spots were recognized by the antiserum raised against tomato fruit AGP holoenzyme, only three of them reacted strongly with antiserum raised against the potato tuber AGP large subunit, and the other two spots (with lower molecular weights) reacted specifically with antisera raised against spinach leaf AGP holoenzyme and the potato tuber AGP small subunit. The results suggest the existence of at least three isoforms of the AGP large subunit and two isoforms of the small subunit in tomato fruit in vivo. The native molecular mass of the enzyme determined by gel filtration was 220 +/- 10 kD, indicating a tetrameric structure for AGP from tomato fruit. The purified enzyme is very sensitive to 3-phosphoglycerate/inorganic phosphate regulation.     

An Antisense Pectin Methylesterase Gene Alters Pectin Chemistry and Soluble Solids in Tomato Fruit

Pectin methylesterase (PME, EC 3.1.11) demethoxylates pectins and is believed to be involved in degradation of pectic cell wall components by polygalacturonase in ripening tomato fruit. We have introduced antisense and sense chimeric PME genes into tomato to elucidate the role of PME in fruit development and ripening. Fruits from transgenic plants expressing high levels of antisense PME RNA showed <10% of wild-type PME enzyme activity and undetectable levels of PME protein and mRNA. Lower PME enzyme activity in fruits from transgenic plants was associated with an increased molecular weight and methylesterification of pectins and decreased levels of total and chelator soluble polyuronides in cell walls. The fruits of transgenic plants also contained higher levels of soluble solids than wild-type fruits. This trait was maintained in subsequent generations and segregated in normal Mendelian fashion with the antisense PME gene. These results indicate that reduction in PME enzyme activity in ripening tomato fruits had a marked influence on fruit pectin metabolism and increased the soluble solids content of fruits, but did not interfere with the ripening process.     

Cell wall-degrading enzymes and pectin solubility and depolymerization in immature and ripe watermelon (Citrullus lanatus) fruit in response to exogenous ethylene

Watermelon fruit have been shown to be extremely sensitive to exogenous ethylene, exhibiting acute symptoms of whole-fruit softening and placental-tissue water-soaking following short periods of exposure to the gas. This study addressed the firmness, specific activities of cell wall hydrolases, and solubility and molecular mass properties of polyuronides in placental tissue in response to treatment of intact fruit with ethylene. Watermelon fruit were harvested at the immature and full-ripe stages and exposed to 50 µl l⁻¹ ethylene for 6 days at 20°C. The firmness of placental tissue from ethylene-treated ripe and immature fruit decreased nearly 80% during 6 days of ethylene exposure, whereas the firmness of placental tissue from fruit maintained in air remained relatively constant up to day 3 and then decreased slightly (12%) during the following 3 days of storage. Although ethylene treatment in some cases influenced the levels of extractable placental-tissue polygalacturonase (EC 3.2.1.15), pectinmethylesterase (EC 3.2.1.11), and α -(EC 3.2.1.22) and β -galactosidase (EC 3.2.1.23) specific activities, these effects were not observed for fruit of both developmental stages and appeared not to be directly involved in the water–soaking syndrome. Symptoms of water-soaking were accompanied by increases in the levels of water- and CDTA ( trans -1,2-cyclohexanediamine- N,N,N',N' -tetraacetic acid)-soluble polyuronides and significant molecular mass downshifts in polyuronides in both immature and ripe watermelon fruit. Polyuronide depolymerization in ethylene-treated ripe fruit was extensive. The parallel trends of enzyme changes in ethylene- compared with air-treated fruit indicate that extractable enzyme levels are not associated with development of the water-soaking disorder. The potential involvement of membrane dysfunction in the water-soaking phenomenon is discussed.     

Purification and catalytic properties of polygalacturonase isoforms from ripe avocado (Persea americana) fruit mesocarp

Endo-polygalacturonase (PG; EC 3.2.1.15) was recovered from the cell walls of avocado mesocarp ( Persea americana Mill cv. Lula) tissue and purified by sequential ion exchange and gel permeation chromatography. Two isoforms (S-I and S-II) were recovered, exhibiting molecular masses of about 41 kD on size exclusion media and about 48 (S-I) and 46 (S-II) kDa as determined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Both isoforms exhibited maximum activity at pH 6.0 against polygalacturonic acid (PGA) and hydrolyzed PGA of about 180 kDa to polymers of about 4 kDa. The catalytic activity of the 48-kDa isoform against PGA was slightly higher than that of the 46-kDa isoform. The purified PGs catalyzed significant molecular mass downshifts in the polyuronides of pre-ripe avocados; however, the capacity of the enzymes to solubilize polyuronides from cell walls of pre-ripe fruit was limited.     

Polyuronide solubilization during ripening of normal and mutant tomato fruit

The amount and molecular size of soluble polyuronide extractable from ripening tomatoes is markedly affected by residual enzyme activity. The efficacy of phenol-acetic acid-water treatment to remove this residual activity is demonstrated. Data obtained using treated wall preparations confirms that there is an increase in soluble polyuronide during normal ripening and that this also occurs in the ‘Never-ripe’ mutant, and to a lesser degree in the ‘ripening-inhibitor’ mutant. However, changes in the molecular size of this polyuronide during normal ripening were not as extensive as previously reported and few changes were apparent in either of the mutants.Measurements were also made of polygalacturonase (EC 3.2.1.15) and pectinesterase (EC 3.1.1.11) activity during ripening. The level of polygalacturonase activity does not appear to correlate with the amount of soluble polyuronide released, but may be related to the extent of depolymerisation. No relationship was apparent between the level of pectinesterase and either soluble polyuronide released or depolymerization.     

Endo-1,4-[beta]-Glucanase,Xyloglucanase, and Xyloglucan Endo-Transglycosylase Activities Versus Potential Substrates in Ripening Tomatoes

In ripening fruits of tomato (Lycopersicon esculentum L. var 83-G-38), the amounts of cellulose and xyloglucan (XG) remained constant during tissue softening, but the relative molecular weight (Mr) of XG decreased markedly and the Mr of cellulose declined slightly. These changes could have been due to activities of non-specific endo-1,4-[beta]-glucanases and/or buffer-soluble XG endo-transglycosylase, both of which increased when tissue firmness declined most rapidly. Tomato extracts also reduced the viscosity of XG solutions, especially in the presence of added XG oligosac-charides. This depolymerizing (XGase) capacity differed from [beta]-glucanase and XG transglycosylase activity (a) by being almost entirely buffer insoluble, and (b) by declining precipitously during fruit softening. Although it disappeared from ripe fruit, XGase may have functioned in promoting wall loosening at earlier stages of fruit development when its activity was highest. By contrast, during aging of fruit in the ripening-inhibited mutant rin there was no change in Mr of XG or cellulose, and activities of [beta]-glucanases and XG transglycosylase were lower than in wild-type tomato. Nevertheless, some softening of the fruit did take place over time and XG amounts declined, possibly because high XGase activity was maintained in the mutant, unlike in wild-type fruit.     

Modification of expansin protein abundance in tomato fruit alters softening and cell wall polymer metabolism during ripening

The role of the ripening-specific expansin Exp1 protein in fruit softening and cell wall metabolism was investigated by suppression and overexpression of Exp1 in transgenic tomato plants. Fruit in which Exp1 protein accumulation was suppressed to 3% that of wild-type levels were firmer than controls throughout ripening. Suppression of Exp1 protein also substantially inhibited polyuronide depolymerization late in ripening but did not prevent the breakdown of structurally important hemicelluloses, a major contributor to softening. In contrast, fruit overexpressing high levels of recombinant Exp1 protein were much softer than controls, even in mature green fruit before ripening commenced. This softening was correlated with the precocious and extensive depolymerization of structural hemicelluloses, whereas polyuronide depolymerization was not altered. These data are consistent with there being at least three components to fruit softening and textural changes. One component is a relaxation of the wall directly mediated by Exp1, which indirectly limits part of a second component due to polyuronide depolymerization late in ripening, perhaps by controlling access of a pectinase to its substrate. The third component is caused by depolymerization of hemicelluloses, which occurs independently of or requires only very small amounts of Exp1 protein.     

Cell wall modification in 1-methylcyclopropene-treated post-climacteric fresh-cut and intact papaya fruit

Papaya is a climacteric fruit in which ripening is greatly regulated by ethylene often associated with stress responses such as wounding. The changes in cell wall compositions in papaya fruit at an advanced stage of ripening under stress conditions including chilling temperature of 5°C and wounding employed as fresh-cut and how these changes were affected by an ethylene action inhibitor of 1-methylcyclopropopene (1-MCP) were examined in the study. The recovery of ethanol-insoluble solids, total soluble sugars, water-soluble polyuronides, neutral hemicelluloses, and neutral sugars of rhamnose, arabinose, mannose and glucose were not affected by 1-MCP or fresh-cut processing. The fresh-cut processing, however, caused a higher loss of total polyuronides and the neutral sugar galactose while increasing the recovery of chelator-soluble polyuronides. Few significant differences due to 1-MCP application were recorded in the recoveries of alkali-soluble polyuronides, hemicellulosic polyuronides extracted with 4% KOH, and the neutral sugar xylose. Modifications of cell wall polyuronides and hemicelluloses in ripe fresh-cut papaya fruit exhibited mostly similar patterns to those in intact ripe papaya fruit under the chilling temperature of 5°C while minimally affected by 1-MCP.     

Galactose loss and fruit ripening: high-molecular-weight arabinogalactans in the pectic polysaccharides of fruit cell walls

Cell wall material (CWM) was prepared from nine fruit species at two ripening stages (unripe and ripe) and extracted sequentially with 0.05 M trans-1,2-diaminocyclohexane-N,N,N′,N′-tetraacetic acid (CDTA), 0.05 M Na₂CO₃ and 4 M KOH. Each solubilised fraction and the CWM-residue remaining after 4 M KOH extraction was analysed for non-cellulosic sugar composition. A common pattern of distribution for polyuronide and pectin-associated neutral sugar was observed for all unripe fruit. Most polyuronide was extracted in the CDTA/Na₂CO₃ fractions while 70–93% of the neutral sugar was located on pectic polysaccharides in the 4 M KOH-soluble and CWM-residue fractions. During ripening, most of the galactose was lost from pectic polysaccharides in the CWM-residue. Partial solubilisation of these polysaccharides was achieved by treating the CWM-residue with endopolygalacturonase. The solubilised polysaccharides were separated into two fractions by ion-exchange chromatography. One of these contained polysaccharides with average molecular weights of 400 kDa or larger and consisted of between 70 and 90% arabinogalactan. The galactosyl residues were 80–90% β-1→4 linked, indicating largely unbranched side-chains. The arabinosyl residues were distributed among terminal, 3-, 5-, 2,5-, and 2,3,5-linked residues, indicating a highly ramified structure. The results are discussed with regard to the relationship between pectin solubilisation and galactose loss and their respective contribution to fruit softening. Received: 28 January 1997 / Accepted: 11 March 1997