Plant Cell, Vol. 11, 2203-2216, November 1999, Copyright © 1999, American Society of Plant Physiologists
Modification of Expansin Protein Abundance in Tomato Fruit Alters Softening and Cell Wall Polymer Metabolism during Ripening
David A. Brummell1,a,
Mark H. Harpster1,a,
Pedro M. Civello3,b,
Joseph M. Palysc,
Alan B. Bennettb, and
Pamela Dunsmuira
a DNA Plant Technology, 6701 San Pablo Avenue, Oakland, California 94608-1239
b Mann Laboratory, Department of Vegetable Crops, University of California, Davis, California 95616
c Seminis Vegetable Seeds, 37437 State Highway 16, Woodland, California 95695
Correspondence to:
David A. Brummell, brummell{at}dnap.com (E-mail), 510-547-2817 (fax)
 |
ABSTRACT |
|---|
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.
|
|
INTRODUCTION |
|---|
Plant cell growth is constrained by the cell wall, whose rigid cellulose microfibrils are held together by three structurally independent but interacting matrices (Carpita and Gibeaut 1993 
). The matrix glycans, or hemicelluloses, are polysaccharides composed of neutral sugars, a major constituent of which in dicotyledonous species is xyloglucan. Xyloglucan molecules coat and cross-link cellulose microfibrils and, by extensive noncovalent bonding, anchor the microfibrils relative to one another. The xyloglucan–cellulose framework is itself embedded in a pectin matrix composed of polyuronide molecules, together with a domain of structural glycoproteins. Precisely how cell wall components and their intermolecular interactions are changed to allow wall loosening during growth or remodeling during cell development is not known, but a group of highly conserved cell wall proteins called expansins have been implicated in many aspects (Cosgrove 1998 ).
Expansins are encoded by large multigene families (Shcherban et al. 1995 ; Cho and Kende 1997 ; Cosgrove et al. 1997 ; Brummell et al. 1999a ), and the expression of expansin mRNA and protein is correlated with growth in many tissues of the plant, including hypocotyls (McQueen-Mason et al. 1992 ), coleoptiles (Li et al. 1993 ), internodes (Cho and Kende 1997 ), leaves (Keller and Cosgrove 1995 ), roots (Wu et al. 1996 ), and green fruit (Brummell et al. 1999a ). However, expansins also may play a role in the cell wall modification involved in other aspects of plant development. In tomato, expansin Exp1 mRNA is expressed at high levels specifically during fruit ripening (Rose et al. 1997 ) and is the major expansin gene family member expressed at this time (Brummell et al. 1999a ). Ripening occurs toward the end of fruit development after fruit growth is complete (Gillaspy et al. 1993 ) and is a complex developmental program that includes changes in color, aroma, metabolism of sugars and organic acids, texture, and softening (Brady 1987 ). These latter two processes are believed to be due to alterations in turgor and cell wall metabolism, which in tomato includes the extensive depolymerization of both pectins and hemicelluloses and slight depolymerization of cellulose (Huber and O'Donoghue 1993 ; Maclachlan and Brady 1994 ).
Little is known about the mechanism of expansin action in cell walls in vivo. Expansins lack significant hydrolase or transglycosylase activity (McQueen-Mason et al. 1992 , McQueen-Mason et al. 1993 ; McQueen-Mason and Cosgrove 1995 ) but share structural features with polysaccharide binding proteins (Shcherban et al. 1995 ) and have been hypothesized to act at the matrix–microfibril interface (McQueen-Mason and Cosgrove 1995 ). In vitro, active expansin protein causes cell wall loosening in isolated cell walls maintained under tension (McQueen-Mason et al. 1992 ; Li et al. 1993 ) and brings about mechanical weakening of paper (McQueen-Mason and Cosgrove 1994 ), which is consistent with expansins acting through a disruption of noncovalent interactions, such as hydrogen bonding. The goal of this project was to suppress and overexpress the Exp1 protein in transgenic tomato plants, first to investigate the contribution of Exp1 protein to the fruit softening process during ripening, and second to gain some insight into the role of expansin protein in fruit cell wall metabolism in vivo.
|
|
RESULTS |
|---|
Suppression and Overexpression of Exp1 in Tomato Plants
Two constructs containing the cDNA of the tomato fruit ripening–specific Exp1 gene in the sense orientation driven by the constitutive cauliflower mosaic virus 35S promoter were assembled (Figure 1A) and used to transform tomato plants. One of these constructs contained the full-length coding region of the cDNA, and the other contained a truncated out-of-frame version in which the region encoding the first 77 amino acids of the deduced Exp1 protein was deleted. Transformation with the full-length construct was intended to result in both silencing of the endogenous Exp1 gene by sense suppression (Napoli et al. 1990 ) and overexpression of functional Exp1 protein, whereas the transcript of the truncated construct could not produce functional Exp1 protein and was intended to result only in sense suppression of the endogenous gene. Plants transformed with the full-length coding Exp1 construct were designated by the prefix 17, whereas the prefix 18 was used to designate plants transformed with the truncated Exp1 construct.
View larger version (61K):
[in this window]
[in a new window]
|
Figure 1.
Constructs and Screening of Primary Transformants.
(A) Schematic representation of Exp1 constructs. For the full-length coding construct (17-series plants), an NcoI site was introduced at the first Met of the Exp1 cDNA by site-directed mutagenesis, before a translationally in-frame fusion to the cauliflower mosaic virus 35S promoter with Cab22L leader. For the out-of-frame truncated construct (18-series plants), an NcoI site was introduced 233 bp downstream of the first Met. nos, nopaline synthase.
(B) Exp1 mRNA abundance in leaves and fruit of plants transformed with the full-length construct (17 series).
(C) Exp1 mRNA abundance in leaves and fruit of plants transformed with the truncated construct (18 series).
In (B) and (C), the three lines selected for further study are indicated with asterisks. WT, wild type.
|
|
Young leaves and red ripe fruit were collected from 136 primary transformants of the 17-series plants and from 49 primary transformants of the 18-series plants. Exp1 mRNA abundance in these tissues was determined by RNA gel blot analysis, as shown for some representative samples in Figure 1B and Figure 1C. Because the endogenous Exp1 gene is not expressed in leaves (Rose et al. 1997 ), Exp1 transcript detected in leaves results from expression of the transgene. In fruit from plants transformed with the full-length coding construct (Figure 1B, bottom), transcripts derived from the endogenous Exp1 gene or from the Exp1 transgene could not be separated by size, but because both transcripts may encode functional Exp1 protein, total Exp1 transcript abundance was determined. In fruit from plants transformed with the truncated construct (Figure 1C, bottom), the transcript derived from the truncated transgene is slightly smaller than mRNA derived from the endogenous Exp1 gene, and the abundance of only the latter was determined. A wide range of Exp1 expression levels in fruit was found, with 8% of the 17-series plants and 18% of the 18-series plants showing suppression of Exp1 mRNA to 10% or less than that of wild-type levels. It was noted that in general, silencing of Exp1 transgene expression in the leaf was correlated with silencing of endogenous Exp1 expression in the fruit; conversely, high expression of the transgene in the leaf was correlated with high endogenous and transgene expression in the fruit. However, in ~10% of the population, high expression in the leaf accompanied low expression in the fruit (e.g., plant number 17-79), or vice versa.
Several lines showing a high level of suppression or overexpression were selected for further study. Segregation analysis (data not shown) was used to determine that strongly suppressed plants 17-42 and 18-40, and overexpressing plant 17-87 possessed single-locus insertions of the selectable marker gene. The T-DNA insertions of these plants were mapped by DNA gel blot analysis. Figure 2 shows that the T-DNA insertions of suppressants 17-42 and 18-40 produced inverted repeats of the transgene, the presence of which frequently has been correlated with post-transcriptional gene silencing (Stam et al. 1997 ; Waterhouse et al. 1998 ). In both lines, the inverted T-DNA repeats were stabilized by asymmetric deletions, which for line 17-42 was of ~900 bp and included the right borders of both T-DNA elements, and for line 18-40 was of ~300 bp, which included the right border of one element. Line 18-40 also had deletions of both T-DNA left borders and part of the selectable marker gene of one element. The overexpressing line 17-87 possessed a single, complete insertion of the T-DNA.
View larger version (29K):
[in this window]
[in a new window]
|
Figure 2.
Schematic Representation of T-DNA Insertions.
T-DNA insertions of three single locus transformants were mapped by DNA gel blot analysis. Transformants 17-42 and 18-40 were suppressants, and transformant 17-87 was an overexpresser. The Gp2:surB chlorsulfuron resistance selectable marker gene is shown as shaded boxes, the 35S promoter as striped boxes, the Exp1 transgene coding region as black boxes, and the nopaline synthase transcription terminator as white boxes. Arrows beneath each gene indicate direction of transcription. LB, T-DNA left border; RB, T-DNA right border.
|
|
For the above-mentioned three transgenic lines, Exp1 mRNA and protein abundance were determined during fruit ripening in the succeeding T2 generation. Figure 3A shows that in azygous control plants of each line, from which the Exp1 transgene had been eliminated by segregation, Exp1 mRNA was not present at the mature green 2 stage but had appeared by the turning stage of ripening, reaching maximum accumulation in red ripe fruit. In homozygous fruit of lines 17-42 and 18-40, Exp1 mRNA accumulation was strongly suppressed throughout ripening, and in homozygous fruit of line 17-87, it was substantially increased. At the red ripe stage, Exp1 mRNA abundance was suppressed to ~5% that of the wild type in line 17-42, ~1% that of the wild type in line 18-40, and increased to ~500% of the wild-type levels in line 17-87. The mRNA accumulation of three related expansin genes expressed predominantly in green fruit, Exp3, Exp4, and Exp5 (Brummell et al. 1999a ), was not altered in Exp1-suppressed fruit (data not shown).
View larger version (60K):
[in this window]
[in a new window]
|
Figure 3.
RNA Gel Blot and Protein Immunoblot Analyses of Exp1 in Azygous and Homozygous Fruit during Ripening.
(A) Total RNA (10 µg per lane) isolated from fruit of the indicated genotype and developmental stage was separated electrophoretically, blotted to a nylon membrane, and Exp1 mRNA detected by hybridizing with radiolabeled Exp1 cDNA. Total cell wall protein (100 µg per lane) prepared from similar fruit tissue was separated by SDS-PAGE, blotted to a polyvinylidene fluoride membrane, and Exp1 protein detected by reaction with Exp1 antiserum. Fruit from azygous plants are shown on the left, and fruit from homozygous plants are on the right. Developmental stages are as follows: MG2, mature green stage 2; TU, turning; PK, pink; LR, light red; RR, red ripe; OR, overripe.
(B) Exp1 protein content in cell walls of red ripe fruit of suppressed and overexpressing lines was estimated relative to wild type (WT) by immunoblot analysis. Total cell wall protein was 100 µg per lane or a dilution of 0.3, 0.1, or 0.03 of 100 µg as indicated. Methods are as given in (A). Gel at the top shows a dilution series of the wild type versus suppressants 17-42 and 18-40. Gel at the bottom shows a dilution series of overexpresser 17-87 versus the wild type.
|
|
Exp1 protein also was absent in cell walls of control fruit at the mature green 2 stage, although a cross-reacting band of slightly higher molecular mass was visible. Exp1 protein accumulation correlated with ripening in cell walls of control fruit, achieving highest levels at the red ripe or overripe stage. Exp1 protein accumulation was strongly suppressed in lines 17-42 and 18-40, although a cross-reacting band of slightly higher molecular mass was seen in some lanes and strongly increased in line 17-87. The relative abundance of Exp1 protein in red ripe fruit was estimated by comparing dilutions of cell wall proteins on an immunoblot. Figure 3B shows that Exp1 protein was present at ~5% of the wild-type levels in line 17-42, 3% of the wild-type levels in line 18-40, and 300% of the wild-type levels in line 17-87. Note that in line 17-87, because overexpression of Exp1 was driven by the 35S promoter, constitutive high-level expression of Exp1 mRNA and protein was detected during fruit development, including at the mature green 2 stage before the onset of ripening.
Suppression of Exp1 Is Correlated with Increased Fruit Firmness and Overexpression with Enhanced Softening
Figure 4 shows the changes in fruit firmness during ripening in T2 generation azygous controls and transgenic lines in which Exp1 mRNA and protein levels were suppressed or overexpressed. In controls of all three lines, fruit firmness declined by approximately one-third between harvest at the midmature green stage and 3 days later. The decline in fruit firmness and the accumulation of Exp1 mRNA both begin late in the mature green stage, immediately before the beginning of ripening (Maclachlan and Brady 1994 ; Brummell et al. 1999a ). After this, fruit firmness decreased progressively during ripening, and from the breaker stage (defined as the first visible exterior red color at the blossom end of the fruit) declined by approximately one-half at the red ripe stage (6 days postbreaker) and by two-thirds at the overripe stage (12 days postbreaker). All three control azygous lines showed very similar softening behavior during ripening.
View larger version (31K):
[in this window]
[in a new window]
|
Figure 4.
Effect of Exp1 Suppression and Overexpression on Fruit Firmness during Ripening.
A minimum of 30 fruit was harvested from azygous control plants (black bars) and from homozygous plants (striped bars) at the mature green stage and ripened off the vine. Fruit compressibility was measured at the indicated intervals as the force required to compress the pericarp by a fixed distance of 0.5 mm (three readings per fruit per time point). Bars represent standard errors, and asterisks indicate that homozygotes were significantly different (P < 0.05) from azygous controls. Developmental stages are as follows: MG, mature green; MG+3d, late mature green; BR, breaker; PK, pink; LR, light red; RR, red ripe; OR, overripe.
(A) Firmness of suppressed line 17-42 relative to azygous controls.
(B) Firmness of suppressed line 18-40 relative to azygous controls.
(C) Firmness of overexpressing line 17-87 relative to azygous controls.
|
|
In line 17-42, Figure 4A shows that suppression of Exp1 protein accumulation to ~5% that of wild-type levels resulted in an increase in fruit firmness early in ripening but that this was no longer apparent from the light red ripening stage onwards. In line 18-40, however, where accumulation of Exp1 protein was reduced to lower levels than in line 17-42 (to ~3% of wild-type levels), Figure 4B shows that firmness was increased at all ripening stages, including the overripe stage. The magnitude of the increase in firmness above azygous controls was similar at each time point, but due to the softening occurring during ripening, the relative enhancement increased from 13% firmer than controls at breaker to 23% firmer at overripe.
Figure 4C shows that constitutive overexpression of recombinant Exp1 protein throughout fruit development, including at the mature green stage, resulted in substantially enhanced fruit compressibility at all stages of fruit development and ripening, particularly in mature green fruit. Mature green fruit of line 17-87 was ~30% softer than controls, and although the magnitude of the difference between homozygous overexpressers and controls declined as ripening progressed, it remained evident even in overripe fruit.
The softening behavior of transgenic T2 fruit ripened on the vine and harvested at the same stages as above showed similar findings to fruit ripened off the vine. Examination of fruit from the succeeding T3 generation also gave comparable results, confirming that the phenotype segregated with the transgene. No difference in ripening rate between transgenic and azygous fruit was noted in fruit ripened either on or off the vine (data not shown).
The average size of Exp1-suppressed fruit was not noticeably different from controls, but fruit overexpressing Exp1 were on average approximately one-half the size of controls. In addition, the pericarp tissue possessed a rubbery texture, suggesting that the overexpression of Exp1 protein affected not only fruit growth but also the properties of the fruit cell wall. The height of plants constitutively overexpressing Exp1 was not significantly reduced, implying that Exp1 may not affect cell expansion, although leaves were slightly smaller and showed some epinasty.
Effect of Modified Exp1 Levels on Fruit Cell Wall Pectins
To investigate the cell wall changes that may underly changes in fruit softening, we examined the depolymerization of cell wall polymers during ripening by using size exclusion chromatography. Cell wall extracts were prepared from control azygous and homozygous fruit of suppressed line 18-40 and overexpresser line 17-87 at four ripening stages. Pectins were extracted in two cell wall fractions: a trans-1,2-diaminocyclohexane-N,N,N',N'-tetraacetic acid (CDTA)–extractable fraction enriched in ionically bound polyuronides, and a Na2CO3–extractable fraction enriched in covalently bound polyuronides. The amounts of pectins solubilized by these two extracts changed with ripening essentially as described previously (Brummell and Labavitch 1997 ), and no significant effects of Exp1 suppression or overexpression on pectin extractability were observed (data not shown).
Figure 5A shows that in control fruit at the mature green stage, CDTA-extractable polyuronides eluted as a major peak in or close to the void (Vo) of the column (molecular mass of 20 x 106 D or greater). During ripening (Figure 5B to 5D), the amount of polyuronide in this peak declined, concomitant with the increasing accumulation of lower molecular mass species nearer the total volume (Vt) of the column (100 kD). However, Figure 5A to 5D show that in fruit suppressed in Exp1 protein accumulation, polyuronide depolymerization was substantially arrested later in ripening. At the red ripe and overripe stages, the largest proportion of the polyuronides eluted between the void and fraction 45, whereas in controls, the bulk of the polyuronides eluted between fraction 50 and the Vt of the column. Although it is not possible to calibrate columns of this separation range due to the lack of available standards, the difference in polyuronide degree of polymerization between control and Exp1-suppressed lines was substantial.
View larger version (19K):
[in this window]
[in a new window]
|
Figure 5.
Size Distribution of CDTA-Extractable Polyuronides from Fruit of Suppressed 18-40 Relative to Controls.
Polyuronides extracted from cell walls of azygous control (solid lines) or homozygous line 18-40 (dashed lines) fruit at four ripening stages were subjected to size exclusion chromatography on a column of Sepharose CL-2B, and column fractions were assayed for uronic acid content. The void (Vo) and total volume (Vt) of the column are marked.
(A) Polyuronides from control and line 18-40 at the mature green 2 stage.
(B) Polyuronides from control and line 18-40 at the pink stage.
(C) Polyuronides from control and line 18-40 at the red ripe stage.
(D) Polyuronides from control and line 18-40 at the overripe stage.
|
|
In contrast, Figure 6A to 6D show that when Exp1 protein was overexpressed, depolymerization of CDTA-extractable polyuronides was indistinguishable from controls at all fruit development stages, including mature green. This suggests that the reduction of cell wall polyuronide depolymerization seen in Exp1-suppressed fruit is most likely due to indirect effects of Exp1 protein on cell wall polyuronide metabolism.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 6.
Size Distribution of CDTA-Extractable Polyuronides from Fruit of Overexpresser 17-87 Relative to Controls.
Polyuronides extracted from cell walls of azygous control (solid lines) or homozygous line 17-87 (dashed lines) fruit at four ripening stages were subjected to size exclusion chromatography on a column of Sepharose CL-2B, and column fractions were assayed for uronic acid content. The void (Vo) and total volume (Vt) of the column are marked.
(A) Polyuronides from control and line 17-87 at the mature green 2 stage.
(B) Polyuronides from control and line 17-87 at the pink stage.
(C) Polyuronides from control and line 17-87 at the red ripe stage.
(D) Polyuronides from control and line 17-87 at the overripe stage.
|
|
Polyuronides extractable in Na2CO3 showed an elution profile quite different from those extracted in CDTA. Figure 7A shows that these polyuronides formed a predominant peak of relatively low molecular mass with a much lower abundance of high molecular mass species. During ripening, no substantial differences to this molecular mass profile were observed, and no effect of suppression or overexpression of Exp1 protein was detected (data not shown). However, this fraction is deesterified by the alkaline Na2CO3 used for extraction, and it is possible that ripening-related changes in intermolecular ester links are masked by this treatment.
View larger version (22K):
[in this window]
[in a new window]
|
Figure 7.
Size Distribution of Na2CO3–Extractable Polyuronides and 4% KOH-Extractable Hemicelluloses from Fruit Cell Walls.
(A) Shown are a representative molecular mass profile of Na2CO3–extractable cell wall polyuronides subjected to size exclusion chromatography on a column of Sepharose CL-2B and column fractions assayed for uronic acid content.
(B) Shown are representative molecular mass profiles of 4% KOH-extractable cell wall hemicelluloses subjected to size exclusion chromatography on a column of Sepharose CL-6B and column fractions assayed for contents of neutral sugars (thick line) and xyloglucan (thin line). Elution points of linear dextran standards are indicated at top in kilodaltons.
The void (Vo) and total volume (Vt) of the columns are indicated.
|
|
Effect of Modified Exp1 Levels on Fruit Cell Wall Hemicelluloses
Cell wall hemicelluloses were extracted from the depectinated cell wall residues by sequential treatment with 4% (w/v) and 24% (w/v) KOH containing 0.1% (w/v) NaBH4. The 4% KOH treatment extracted approximately one-third of the KOH-extractable neutral sugar polysaccharides, and the 24% KOH treatment extracted two-thirds. These proportions were not affected either by ripening stage or by suppression or overexpression of Exp1 protein. Xyloglucan was found predominantly in the latter extract.
Figure 7B shows that the 4% KOH-extractable hemicelluloses eluted from the column with a molecular mass profile composed of four main peaks, in the void (equivalent to dextran standards of 1000 kD or greater), fraction 47 (200 kD), fraction 58 (70 kD), and fraction 62 (40 kD), of which the most prominant peak was at 70 kD. Xyloglucan was present at low levels coincident with the 200-kD peak. No substantial changes to these molecular mass profiles occurred during ripening, and no effect of either suppression or overexpression of Exp1 protein was observed (data not shown).
For fruit suppressed in Exp1 protein and for corresponding controls, molecular mass profiles of 24% KOH-extractable hemicelluloses were determined for both all neutral sugar-containing polysaccharides (Figure 8A to 8D) and xyloglucan (Figure 8E to 8H). Figure 8A shows that in mature green fruit of controls, total hemicellulosic polysaccharides eluted from the column with a molecular mass profile composed of four major peaks. The two predominant peaks overlapped and possessed maxima in the void of the column (1000 kD or greater) and in fraction 38 (350 kD). The two minor peaks possessed maxima of ~300 kD (fraction 42) and 70 kD (fraction 58). During ripening in control fruit (Figure 8A to 8D), there was a loss of hemicelluloses eluting in the void and a reduction in the amount of material eluting in the 350-kD peak, coupled with a concomitant increase in the amount of material in the 300-kD peak, which also showed depolymerization during ripening, and in overripe fruit had a maximum at 200 kD (fraction 46). The hemicellulose peak at ~70 kD showed little change during ripening.
View larger version (32K):
[in this window]
[in a new window]
|
Figure 8.
Size Distribution of Total Hemicelluloses and Xyloglucan from Fruit of Suppressant 18-40 Relative to Controls.
Hemicelluloses extracted from cell walls of azygous control (solid lines) or homozygous line 18-40 (dashed lines) fruit at four ripening stages were subjected to size exclusion chromatography on a column of Sepharose CL-6B, and column fractions were assayed for contents of neutral sugar (total hemicelluloses) ([A] to [D]) and xyloglucan ([E] to [H]). Void (Vo) and total volume (Vt) of the column and elution points of linear dextran standards are indicated in kilodaltons at top.
(A) to (D) Hemicelluloses from control and line 18-40 at the mature green 2 (A), pink (B), red ripe (C), and overripe (D) ripening stages.
(E) to (H) Xyloglucan from control and line 18-40 at the mature green 2 (E), pink (F), red ripe (G), and overripe (H) ripening stages.
|
|
Analysis of xyloglucan molecular mass profiles in controls (Figure 8E to 8H) showed that in mature green fruit, xyloglucan molecules formed a polydisperse peak with a broad maximum extending from the void of the column to ~350 kD. During ripening, xyloglucan showed substantial depolymerization and became less polydisperse, with the greatest change seen between the pink and red ripe ripening stages, and in overripe fruit became a sharp peak with a maximum at ~300 kD. The changes in the molecular mass profile of xyloglucan occurring during ripening were consistent with the changes seen in the molecular mass profile of total hemicelluloses. However, xyloglucan constitutes at most only one-third of the polysaccharides in this extract, and ripening-related changes in other hemicellulosic polysaccharides also must be occurring.
In fruit suppressed in Exp1 protein, the molecular mass profiles of both total hemicelluloses (Figure 8A to 8D) and xyloglucan (Figure 8E to 8H) were very similar to the control molecular mass profiles at every ripening stage except pink. This probably was due to the difficulties of obtaining exactly matched fruit at the pink ripening stage (polyuronides from these same fruit also were slightly more depolymerized than controls at the pink stage; Figure 5B) rather than to an effect of Exp1 suppression, because the difference was not present at the red ripe and overripe stages. Thus, suppression of Exp1 protein accumulation to 3% of wild-type levels did not detectably affect cell wall hemicellulose molecular mass profiles during ripening, suggesting that the increased firmness of Exp1-suppressed fruit is not due to a modification of hemicellulose depolymerization.
The effect of overexpression of Exp1 protein on hemicellulose depolymerization is shown in Figure 9. During ripening in control fruit, the molecular mass profiles of both total hemicelluloses (Figure 9A to 9D) and xyloglucan (Figure 9E to 9H) were very similar to the control fruit shown in Figure 8. However, at the mature green stage, fruit overexpressing Exp1 (Figure 9A) showed a hemicellulose molecular mass profile more characteristic of control fruit at the overripe stage in that there was little material in the void or 350-kD peaks and there was a predominant peak spanning the 300- to 200-kD size range. The peak equivalent to ~70 kD also was slightly more pronounced than in controls. The molecular mass profile of xyloglucan showed corresponding changes and at the mature green stage (Figure 9E) was also already extensively depolymerized relative to controls. In overexpressing fruit, there was little further change in the molecular mass profiles of either total hemicelluloses or xyloglucan with ripening, whereas the molecular mass profiles of controls became more like those of the Exp1 overexpressers as ripening advanced. Thus, the overexpression of Exp1 protein is correlated with cell wall hemicellulose depolymerization and fruit softening typical of ripe fruit, even in mature green fruit before the commencement of ripening.
View larger version (35K):
[in this window]
[in a new window]
|
Figure 9.
Size Distribution of Total Hemicelluloses and Xyloglucan from Fruit of Overexpresser 17-87 Relative to Controls.
Hemicelluloses extracted from cell walls of azygous control (solid lines) or homozygous line 17-87 (dashed lines) fruit at four ripening stages were subjected to size exclusion chromatography on a column of Sepharose CL-6B, and column fractions were assayed for contents of neutral sugar (total hemicelluloses) ([A] to [D]) and xyloglucan ([E] to [H]). Void (Vo) and total volume (Vt) of the column and elution points of linear dextran standards are indicated in kilodaltons at top.
(A) to (D) Hemicelluloses from control and line 17-87 at the mature green 2 (A), pink (B), red ripe (C), and overripe (D) ripening stages.
(E) to (H) Xyloglucan from control and line 17-87 at the mature green 2 (E), pink (F), red ripe (G), and overripe (H) ripening stages.
|
|
|
|
DISCUSSION |
|---|
The softening of tomato fruit during ripening is accompanied by alterations both in the architecture and physicochemical properties of the cell wall, and in the polymers of which it is composed. The most prominant changes are the solubilization and depolymerization of polyuronides (Huber and O'Donoghue 1993 ; Brummell and Labavitch 1997 ), the depolymerization of hemicelluloses (Tong and Gross 1988 ; Maclachlan and Brady 1994 ), and the loss of galactose (Gross 1984 ). Both the pectins and hemicelluloses are composed of a diverse range of polysaccharides; the hemicelluloses, including xyloglucan, glucomannans, xylans, arabinans, and galactans; and the pectins, homogalacturonans, and smaller amounts of heterogeneous rhamnogalacturonans (Seymour et al. 1990 ; Carpita and Gibeaut 1993 ). During ripening, coordinated increases occur in the mRNA abundance and activity of a range of enzymes that may contribute to cell wall polymer modification or depolymerization, including endopolygalacturonase (PG; Brady et al. 1982 ; DellaPenna et al. 1986 ), pectin methylesterase (PME; Harriman et al. 1991 ), endo-1,4-ß-glucanase (EGase; Hobson 1968 ; Lashbrook et al. 1994 ), xyloglucan endotransglycosylase (XET; Arrowsmith and de Silva 1995 ), ß-galactosidase (Carey et al. 1995 ; Smith et al. 1998 ), and expansin (Rose et al. 1997 ). How these various enzymes act cooperatively or independently to achieve the precise disassembly of particular cell wall components or complexes during fruit softening is not known.
Role of Exp1 Protein in Pectin Depolymerization
Figure 5 shows that in control fruit, considerable polyuronide depolymerization occurs in a CDTA-extractable pectin fraction during ripening but that this is substantially arrested when Exp1 protein is suppressed to ~3% of wild-type levels. Although pectin disassembly may be brought about by several enzymes, PG is believed to be responsible for a large part of polyuronide depolymerization, and its mRNA, protein, and activity accumulate to very high levels late in the ripening of tomato fruit (Tucker et al. 1980 ; Brady et al. 1982 ; DellaPenna et al. 1986 ; Spiers et al. 1989 ). However, PG-mediated polyuronide depolymerization does not appear to be the primary determinant of fruit softening. Antisense suppression of PG activity to 0.5% of wild-type levels (Sheehy et al. 1988 ; Smith et al. 1988 ) resulted in only a modest reduction of polyuronide depolymerization (Brummell and Labavitch 1997 ) and a very small increase in firmness later in ripening (Kramer et al. 1992 ; Langley et al. 1994 ). Furthermore, expression of a chimeric PG transgene in a nonsoftening mutant tomato fruit that normally lacks this activity resulted in polyuronide solubilization but did not restore softening (Giovannoni et al. 1989 ).
PG catalyzes massive polyuronide degradation in vitro, but in the cell wall its activity appears to be limited by some chemical or physical restriction (Seymour et al. 1987 ; Huber and O'Donoghue 1993 ; Brummell and Labavitch 1997 ). Pectins secreted to the cell wall possess methylester side chains, which are removed by PME as a necessary prerequisite for PG action (Carpita and Gibeaut 1993 ). Also, in the wall, PG may exist in association with the ß subunit, a small protein that could restrict PG activity by binding either to the enzyme or to its substrate (Watson et al. 1994 ). Antisense suppression of PME or ß-subunit protein accumulation reduced or slightly increased polyuronide depolymerization, respectively, but with little effect on fruit softening during ripening (Tieman et al. 1992 ; Tieman and Handa 1994 ; Watson et al. 1994 ).
In contrast, strong suppression of Exp1 protein accumulation in line 18-40 resulted in a moderate increase in fruit firmness throughout ripening (Figure 4B), part of which may be due to reduced pectin breakdown. The reduction of polyuronide depolymerization in Exp1-suppressed fruit was evident late in ripening; in tomato, accumulation of high levels of PG also occurs relatively late in ripening (Tucker et al. 1980 ; Spiers et al. 1989 ), indicating that PG may be the pectinase whose activity is restricted by Exp1 action. Indeed, the reduction of polyuronide depolymerization by suppression of Exp1 was greater than that due to suppression of PG activity itself (Brummell and Labavitch 1997 ) but is presumably an indirect effect. Figure 6A shows that overexpression of Exp1 protein in mature green fruit, where Exp1 is not normally present, does not alter the polyuronide molecular mass profile, suggesting that Exp1 does not itself cause pectin hydrolysis. Later in ripening, when PG appears, Figure 6B to 6D show that pectin depolymerization occurs equivalently in control and overexpressing fruit, and Exp1 overexpression did not significantly increase this process.
Examination of PG mRNA abundance in Exp1-suppressed fruit found that the accumulation of PG mRNA was reduced relative to controls, by three- to fivefold late in ripening (data not shown). However, PG mRNA abundance was still substantial in these fruit, suggesting that although there may be some interaction between Exp1 action and PG protein accumulation, this was not sufficient to cause the observed reduction in polyuronide depolymerization. Overall, the data suggest that the action of Exp1 protein on cell wall structure may be necessary to allow some component of polyuronide disassembly to occur, perhaps by controlling access of PG to its substrate.
Role of Exp1 Protein in Hemicellulose Depolymerization
Figure 7B and Figure 8 show that the 24% KOH-extractable hemicelluloses, which are tightly bound to microfibrils, were generally of higher molecular mass than the less tightly associated 4% KOH-extractable hemicelluloses and contained most of the xyloglucan. Ripening was accompanied by a depolymerization of high molecular mass hemicelluloses, including xyloglucan (Sakurai and Nevins 1993 ; Maclachlan and Brady 1994 ), which was limited to polymers tightly bound to cellulose, as previously found (Tong and Gross 1988 ). Tomato fruit contain several enzymic activities that may contribute to hemicellulose depolymerization, notably EGases and XETs. At least two EGases, Cel1 and Cel2, show increases in mRNA abundance coincident with ripening (Lashbrook et al. 1994 ), but antisense suppression of each of these genes separately did not detectably alter fruit softening (Lashbrook et al. 1998 ; Brummell et al. 1999b ). XET activity may be involved specifically in xyloglucan modification, and although total XET activity was much higher in green rather than in red fruit (Faik et al. 1998 ), the mRNA of one XET gene increased in abundance during ripening (Arrowsmith and de Silva 1995 ).
Figure 8 shows that suppression of Exp1 protein accumulation to 3% of wild-type levels does not prevent depolymerization of hemicelluloses during ripening, whereas Figure 9 shows that overexpression of high levels of Exp1 can potentiate hemicellulose depolymerization. The relationship between Exp1 action and hemicellulose depolymerization can be explained by three basic mechanisms. One possibility is that the modification of the cell wall brought about by Exp1 and the depolymerization of hemicelluloses are independent events in vivo. Suppression of Exp1 protein to very low levels thus would be without effect on hemicellulose metabolism. If Exp1 is not normally involved in hemicellulose metabolism, its overexpression in green fruit then may cause the observed hemicellulose depolymerization through some aberrant mechanism. Green fruit possess high levels of hydrolases, including EGases and XETs (Maclachlan and Brady 1994 ; Faik et al. 1998 ), and the presence of abnormally high amounts of Exp1 protein may increase the accessibility of substrate sites in hemicellulose molecules to these nonripening-related enzymes that normally are unable to attack them. The consequence of this would be a degradation of hemicelluloses, similar but perhaps not identical to that occurring during ripening but nevertheless resulting in fruit softening. In green fruit, overexpression of Exp1 to 300% of the levels found in wild-type ripe fruit potentiates hemicellulose breakdown, but it is not known if 3% of wild-type levels can evoke the same effect.
A second possibility is that Exp1 protein is necessary for hemicellulase action, as it is for some pectinase activities (Figure 5). Exp1 may normally function by exposing previously unavailable structural hemicellulose molecules to other degradative enzymes (Rose et al. 1997 ). The observation that constitutive overexpression of Exp1 promotes depolymerization of hemicelluloses in green fruit appears to support this suggestion. However, for this to be the mechanism, 3% of Exp1 protein would need to be sufficient to cause complete hemicellulose availability during ripening, although not enough for 100% of wild-type function in other aspects. For example, the indirect inhibition of polyuronide depolymerization shown in Figure 5 and the increase in fruit firmness shown in Figure 4B show that some components of cell wall metabolism are affected substantially by a 97% suppression of Exp1 protein. Suppression of Exp1 protein abundance to below 3% of wild-type levels, or preferably gene inactivation, is necessary to test this hypothesis. If, however, this mechanism proves to be the case, the regulation by Exp1 of both wall relaxation and the access of particular hemicellulases and pectinases to their substrates would imply a crucial role for this protein in fruit softening.
A third possibility is that hemicellulases are necessary for Exp1 action. Although an interaction at some level should not be excluded, this seems unlikely because, in vitro, expansins can cause extensive and prolonged cell wall loosening in the absence of active cell wall hydrolases (McQueen-Mason et al. 1992 ; McQueen-Mason and Cosgrove 1994 , McQueen-Mason and Cosgrove 1995 ). Thus, the role of Exp1 protein in hemicellulose metabolism remains equivocal, and further data are necessary to discriminate between the first two models. If Exp1 action is the trigger for hemicellulose breakdown during ripening, then hydrolytic activities already present in green fruit before ripening begins may make significant contributions to ripening-related cell wall changes. On the other hand, if Exp1 action and hemicellulose breakdown are independent, the appearance of ripening-related hemicellulases may be the primary determinant of hemicellulose depolymerization.
Expansin, Fruit Cell Wall Metabolism, and Softening
Fruit softening is probably caused by the cumulative effect of a range of modifications occurring in the networks of polymers making up the primary cell wall, all of which contribute in different ways to a loss of firmness and changes in textural qualities. During ripening, fruit softening is associated with depolymerizations of high molecular mass hemicelluloses, including xyloglucan, which are tightly associated with cellulose, and of ionically bound polyuronides. Suppression and overexpression of Exp1 protein have opposite effects on fruit softening, indicating a role for Exp1 in this process, but unexpectedly they also cause indirect changes in different cell wall components.
Overexpression of recombinant Exp1 protein in mature green fruit evokes extensive hemicellulose depolymerization and considerable softening, in the absence of polyuronide depolymerization. Thus, polyuronide depolymerization is not essential for at least one major component of softening. Figure 4C shows that fruit overexpressing Exp1 continue to soften during ripening, even though changes to the hemicellulose molecular mass profile are complete at the mature green stage, implying that mechanisms of softening other than hemicellulose breakdown occur as these fruit ripen. The magnitude of the difference in softness between overexpressing fruit and controls diminishes progressively with ripening, presumably as hemicelluloses of control cell walls also become increasingly depolymerized. In contrast, strong suppression of Exp1 protein accumulation increases fruit firmness to a similar extent throughout ripening (Figure 4B). Suppression of Exp1 protein, however, does not prevent hemicellulose depolymerization, which proceeds at wild-type levels, even though indirect effects on polyuronide disassembly are evident later in ripening. These findings suggest that the increased firmness of suppressed fruit is not due to reduced hemicellulose breakdown but may be due to the suppression of a direct effect of Exp1 protein on cell wall properties throughout ripening and to reduced polyuronide breakdown later in ripening.
Overall, the data are consistent with at least three processes contributing to textural changes and the loss of fruit firmness during ripening. Exp1 protein is responsible for one component, perhaps by a loosening of noncovalent linkages between unidentified polymers at the hemicellulose–microfibril interface, as suggested by McQueen-Mason and Cosgrove 1995 . This is accompanied by the second and major component of softening, consisting of the depolymerization of structurally important hemicelluloses brought about by an independent mechanism that does not require Exp1 protein or for which only very small amounts of Exp1 protein are sufficient. A third aspect of firmness loss is polyuronide depolymerization, the later stages of which are dependent on the Exp1-mediated relaxation of the wall structure necessary to allow PG or other enzymes access to polyuronide substrate sites. This would have the effect of ensuring that cell wall loosening and the complete fruit softening process precede the final component of polyuronide disassembly, which is involved in fruit deterioration. The data also suggest that at least some parts of the softening program occur in sequential steps that are dependent on the completion of previous steps.
|
|
METHODS |
|---|
Production of Transgenic Plants
Using site-directed mutagenesis with mismatched oligonucleotides followed by polymerase chain reaction amplification, we introduced an NcoI restriction site into the Exp1 cDNA at the ATG codon of the translational start and an XbaI site in the 3' untranslated region 51 nucleotides after the translational stop codon. An out-of-frame truncated version of the Exp1 cDNA was produced similarly, with the NcoI site introduced 233 nucleotides into the coding region. Both the in-frame full-length coding and the out-of-frame truncated versions of the mutagenized Exp1 cDNA were ligated into the expression cassette of plasmid pJJ2104 (Harpster et al. 1988 ) between the cauliflower mosaic virus 35S promoter:Cab22L leader and the nopaline synthase 3' transcription terminator (Figure 1A). These constructs were transferred to the binary vector pWTT2179, a derivative of pWTT2084 (Firoozabady et al. 1995 ) in which the surB chlorsulfuron resistance gene was instead driven by the pepper Gp2 promoter (Baden et al. 1997 ). Agrobacterium tumefaciens LBA4404 harboring these constructs was used to transform cotyledons of tomato (Lycopersicon esculentum cv Ailsa Craig) essentially as described previously (Yoder et al. 1988 ). Regenerated transformed seedlings were selected on chlorsulfuron-containing media and rooted and grown to maturity in a greenhouse.
Screening Transgenic Plants by RNA Gel Blot Analysis
RNA was prepared from young leaves and red ripe fruit of primary transformants as described by Dunsmuir et al. 1987 , and aliquots of 10 µg of total RNA were subjected to electophoresis in denaturing gels and then blotted to Duralon-UV membranes (Stratagene, La Jolla, CA). RNA gel blots were hybridized with a 500-bp fragment of the Exp1 cDNA (composed of nucleotides 151 to 648 of the coding sequence) radiolabeled with 
-32P-dCTP, random hexamers, and the Klenow fragment of DNA polymerase I and then washed in 0.1 x SSC (1 x SSC is 0.15 M NaCl and 0.015 M sodium citrate) and 0.1% SDS at 65°C.
Segregation Analysis
Selected primary transformants (T1 generation) were allowed to self-fertilize, and 128 seeds per line of the next (segregating) T2 generation were grown in chlorsulfuron selection to determine segregation patterns of the transgenes. Lines with segregation patterns consistent with a single locus T-DNA insertion were retained. Twelve survivors per line of this chlorsulfuron treatment, that is, hemizygous or homozygous T2 plants, were grown to maturity in a greenhouse. The zygosity of these plants was determined by determining the segregation pattern in the following T3 generation, and homozygous T2 plants were retained. Corresponding azygous plants for each line, from which the transgene had been eliminated by segregation, were identified by polymerase chain reaction and confirmed by 100% chlorsulfuron susceptibility of the seeds. Azygous and homozygous T2 sibling plants were propagated by cuttings and provided the source of all fruit used in this study.
Fruit Material
Fruit were collected from homozygous transgenic plants and corresponding azygous control plants as described above, at particular ripening stages defined by internal anatomy and color (Gonzalez-Bosch et al. 1996 ) or by days after the breaker stage. Breaker stage is the first visible exterior red color at the blossom end of the fruit. The stages harvested were as follows: mature green stage 2 (full-size green fruit, one or two locules liquified), turning (10% red, approximately breaker +0.5 days), pink (50% red, approximately breaker +2 days), light red (100% light red, approximately breaker +4 days), red ripe (breaker +6 days, full red and edibly soft), and overripe (breaker +14 days, very soft with evident deterioration). In wild-type fruit, mature green stage 2 is before the accumulation of Exp1 mRNA, which appears by mature green stage 4 just before breaker (Brummell et al. 1999a ). Three fruit per genotype and ripening stage were collected, locules and seeds were removed, and pooled pericarp material was frozen in liquid nitrogen and then stored at -80°C. Samples of the same fruit material were used for all subsequent RNA, protein, and cell wall analyses. RNA gel blot analysis was conducted as above, and relative mRNA abundances were determined using a PhosphorImager (Molecular Dynamics, Sunnyvale, CA).
Mapping T-DNA Insertions
Genomic DNA was prepared (Greene et al. 1994 ) from young leaves of wild-type and selected transformants, and aliquots of 10 µg were digested with a battery of restriction enzymes. Restriction fragments were separated by electrophoresis on 0.8% agarose gels and blotted to Duralon-UV membranes. Replicate gel blots were hybridized as above with radiolabeled riboprobes corresponding to the Exp1 gene, the chlorsulfuron resistance surB-selectable marker gene, and the right and left T-DNA borders. T-DNA insertions were mapped by comparison to wild type of the number and size of restriction fragments binding to these probes.
Cell Wall Protein Extraction and Protein Gel Blot Analysis
Frozen fruit pericarp tissue was powdered in liquid nitrogen and then homogenized with 2 volumes of homogenization buffer (40 mM sodium phosphate, pH 7.0, 3 mM EDTA, 5 mM DTT, and a cocktail of protease inhibitors). Supernatants containing soluble proteins were discarded, and pellets were resuspended in SDS loading buffer (Laemmli 1970 ), which was supplemented with 5% 2-mercaptoethanol but from which bromophenol blue was omitted. Samples were heated in a boiling water bath for 10 min, mixed, heated for a further 10 min, and centrifuged, and the supernatants containing cell wall proteins were aspirated. This extraction procedure solubilized all immunodetectable Exp1 protein from crude cell wall preparations. Protein contents were quantified relative to dilutions of a BSA standard by spotting dilutions onto 3MM paper (Whatman, Maidstone, UK), washing with acetone, and staining with Coomassie Brilliant Blue R 250.
Exp1 antiserum was a generous gift of J. Rose (University of Georgia, Athens). Antiserum was purified before use as described previously (Harpster et al. 1998 ) by incubating with cell wall protein from immature green fruit, lacking Exp1 expression, which had been covalently coupled to an N-hydroxysuccinimide–activated Sepharose column (Pharmacia, Piscataway, NJ) to form an affinity column. Unbound antiserum was enriched for antibodies specific to Exp1 protein.
Cell wall proteins (100 µg per lane) were separated by SDS-PAGE and electroblotted to Immobilon-P membrane (Millipore, Bedford, MA). Protein gel blots were blocked with nonfat dried milk, incubated with purified Exp1 antiserum at a 1:5000 dilution overnight, washed, reacted with goat anti–rabbit antibody coupled to alkaline phosphatase, and rewashed. Bound antibodies were detected using 5-bromo-4-chloro-3-indolyl phosphate and nitro blue tetrazolium as substrate.
Measurement of Fruit Firmness
Fruit were harvested from homozygous and corresponding azygous control plants at the mature green stage and ripened in air at 20°C. Pericarp firmness was determined using a model TA.XT2 Texture Analyzer (Stable Micro Systems, Godalming, UK) fitted with a flat probe 25 mm in diameter operated at 0.5 mm sec-1, at the following intervals: mature green, mature green +3 days (equivalent to late mature green, i.e., full-size green fruit, all locules liquified, some internal coloration), breaker, breaker +2 days (pink), breaker +4 days (light red), breaker +6 days (red ripe), and breaker +12 days (overripe). At each time point, the force required to compress the pericarp by 0.5 mm was recorded at three different points on the equator of each fruit. Data were analyzed by ANOVA, and means were compared using the Least Significant Difference test at a significance level of 0.05.
Cell Wall Fractionation and Size-Exclusion Chromatography
Acetone-insoluble cell wall material was prepared from frozen fruit pericarp by using the Tris-buffered phenol method (Huber and O'Donoghue 1993 ). Aliquots of 100 mg were sequentially extracted with trans-1,2-diaminocyclohexane-N,N,N',N'-tetraacetic acid (CDTA) and Na2CO3 to extract pectins, as described previously (Brummell and Labavitch 1997 ), and then with 4% and 24% KOH to extract hemicelluloses, as described previously (Maclachlan and Brady 1994 ). Equal amounts (1 mg) of dialyzed CDTA-extractable and Na2CO3–extractable polyuronides were applied to a column of Sepharose CL-2B (96 x 1.5 cm) eluted with 0.2 M ammonium acetate, pH 5.0, and fractions of 2 mL were collected. Uronic acid contents of aliquots of column fractions (0.4 mL) were determined by the method of Blumenkrantz and Asboe-Hansen 1973 . KOH-extractable hemicelluloses were neutralized, precipitated, and treated with -amylase (Maclachlan and Brady 1994 ), and equal amounts (2 mg) were applied to a column of Sepharose CL-6B (97 x 1.5 cm) and eluted with 0.1 M NaOH, and fractions of 2 mL were collected. Column fractions were assayed for neutral sugar content (aliquot of 0.4 mL) by the phenol-sulfuric method (Dubois et al. 1956 ) and for xyloglucan content (aliquot of 0.8 mL) by the iodine binding method (Kooiman 1960 ), an assay that detects polymeric xyloglucan greater than ~20 kD. The CL-6B column was calibrated with linear dextrans of known size, but because these may possess a different conformation to cell wall hemicelluloses, they should be used as an estimate of molecular mass only.
|
|
FOOTNOTES |
|---|
1 These authors contributed equally to this work. 
3 Current address: INTECH, Camino Circunvalacion Laguna Km 6, Casillade Correo 164, 7130 Chascomus, Argentina.
|
|
ACKNOWLEDGMENTS |
|---|
We thank Joss Rose (University of Georgia) for mutagenesis and the generous gift of Exp1 antiserum, Kathy Lee (DNA Plant Technology [DNAP]) for assembly of constructs, Fereshteh Saghaei and Stacey Crocker (Seminis Vegetable Seeds) for assistance with plant transformations, Carol McGugin (DNAP) for segregation analysis, Jay Maddox and Vinnie Pirozzi (DNAP) for maintaining greenhouse plants, Kristen Hadfield and Adnan Sabehat (University of California at Davis) for assistance with fruit collection and firmness measurements, and Ed Ralston and Steve Scofield (DNAP) for comments on the manuscript. We are particularly indebted to Joss Rose for his contribution during the early stages of this project. This work was supported by Seminis Vegetable Seeds.
Received June 23, 1999; accepted September 10, 1999.
|
|
REFERENCES |
|---|
Arrowsmith, D.A., and de Silva, J. (1995) Characterization of two tomato fruit–expressed cDNAs encoding xyloglucan endo-transglycosylase. Plant Mol. Biol. 28:391-403[CrossRef][Web of Science][Medline].
Baden, C., Dunsmuir, P., and Lee, K.Y. (1997). Plant group 2 promoters and uses thereof. U.S. Patent No. 5,608,144. Issued March 4, 1997.
Blumenkrantz, N., and Asboe-Hansen, G. (1973) New method for quantitative determination of uronic acids. Anal. Biochem. 54:484-489[CrossRef][Web of Science][Medline].
Brady, C.J. (1987) Fruit ripening. Annu. Rev. Plant Physiol. 38:155-178[CrossRef][Web of Science].
Brady, C.J., MacAlpine, G., McGlasson, W.B., and Ueda, Y. (1982) Polygalacturonase in tomato fruits and the induction of ripening. Aust. J. Plant Physiol. 9:171-178.
Brummell, D.A., and Labavitch, J.M. (1997) Effect of antisense suppression of endopolygalacturonase activity on polyuronide molecular weight in ripening tomato fruit and in fruit homogenates. Plant Physiol. 115:717-725[Abstract].
Brummell, D.A., Harpster, M.H., and Dunsmuir, P. (1999a) Differential expression of expansin gene family members during growth and ripening of tomato fruit. Plant Mol. Biol. 39:161-169[CrossRef][Web of Science][Medline].
Brummell, D.A., Hall, B.D., and Bennett, A.B. (1999b) Antisense suppression of tomato endo-1,4-ß-glucanase Cel2 mRNA accumulation increases the force required to break fruit abscission zones but does not affect fruit softening. Plant Mol. Biol. 40:615-622[CrossRef][Web of Science][Medline].
Carey, A.T., Holt, K., Picard, S., Wilde, R., Tucker, G.A., Bird, C.R., Schuch, W., and Seymour, G.B. (1995) Tomato exo-(1-4)-ß-D-galactanase. Isolation, changes during ripening in normal and mutant tomato fruit, and characterization of a related cDNA clone. Plant Physiol. 108:1099-1107[Abstract].
Carpita, N.C., and Gibeaut, D.M. (1993) Structural models of primary cell walls in flowering plants: Consistency of molecular structure with the physical properties of the walls during growth. Plant J. 3:1-30[CrossRef][Web of Science][Medline].
Cho, H.-T., and Kende, H. (1997) Expression of expansin genes is correlated with growth in deepwater rice. Plant Cell 9:1661-1671[Abstract].
Cosgrove, D.J. (1998) Cell wall loosening by expansins. Plant Physiol. 118:333-339[Free Full Text].
Cosgrove, D.J., Bedinger, P., and Durachko, D.M. (1997) Group I allergens of grass pollen as cell wall–loosening agents. Proc. Natl. Acad. Sci. USA 94:6559-6564[Abstract/Free Full Text].
DellaPenna, D., Alexander, D.C., and Bennett, A.B. (1986) Molecular cloning of tomato fruit polygalacturonase: Analysis of polygalacturonase mRNA levels during ripening. Proc. Natl. Acad. Sci. USA 83:6420-6424[Abstract/Free Full Text].
Dubois, M., Gilles, K.A., Hamilton, J.K., Rebers, P.A., and Smith, F. (1956) Colorimetric method for determination of sugars and related substances. Anal. Chem. 28:350-356[CrossRef].
Dunsmuir, P., Bond, D., Lee, K., Gidoni, D., and Townsend, J. (1987) The expression of introduced genes in regenerated plants. In Gelvin S., Schilperoort R., eds. Plant Molecular Biology Manual. Dordrecht, The Netherlands, Kluwer Academic Publishers. 45–59.pp.
Faik, A., Desveaux, D., and Maclachlan, G. (1998) Enzymic activities responsible for xyloglucan depolymerization in extracts of developing tomato fruit. Phytochemistry 49:365-376.
Firoozabady, E., Moy, Y., Tucker, W., Robinson, K., and Gutterson, N. (1995) Efficient transformation and regeneration of carnation cultivars using Agrobacterium.. Mol. Breeding 1:283-293.
Gillaspy, G., Ben-David, H., and Gruissem, W. (1993) Fruits: A developmental perspective. Plant Cell 5:1439-1451[Free Full Text].
Giovannoni, J.J., DellaPenna, D., Bennett, A.B., and Fischer, R.L. (1989) Expression of a chimeric polygalacturonase gene in transgenic rin (ripening inhibitor) tomato fruit results in polyuronide degradation but not fruit softening. Plant Cell 1:53-63[Abstract/Free Full Text].
Gonzalez-Bosch, C., Brummell, D.A., and Bennett, A.B. (1996) Differential expression of two endo-1,4-ß-glucanase genes in pericarp and locules of wild-type and mutant tomato fruit. Plant Physiol. 111:1313-1319[Abstract].
Greene, B., Walko, R., and Hake, S. (1994) Mutator insertions in an intron of the maize knotted1 gene result in dominant suppressible mutations. Genetics 138:1275-1285[Abstract].
Gross, K.C. (1984) Fractionation and partial characterization of cell walls from normal and non-ripening mutant tomato fruit. Physiol. Plant. 62:25-32.
Harpster, M.H., Townsend, J.A., Jones, J.D.G., Bedbrook, J., and Dunsmuir, P. (1988) Relative strengths of the 35S cauliflower mosaic virus, 1', 2', and nopaline synthase promoters in transformed tobacco, sugarbeet, and oilseed rape callus tissue. Mol. Gen. Genet. 212:182-190[CrossRef][Medline].
Harpster, M.H., Brummell, D.A., and Dunsmuir, P. (1998) Expression analysis of a ripening-specific, auxin-repressed endo-1,4-ß-glucanase gene in strawberry. Plant Physiol. 118:1307-1316[Abstract/Free Full Text].
Harriman, R.W., Tieman, D.M., and Handa, A.K. (1991) Molecular cloning of tomato pectin methylesterase gene and its expression in Rutgers, ripening inhibitor, nonripening, and Never Ripe tomato fruits. Plant Physiol. 97:80-87[Abstract/Free Full Text].
Hobson, G.E. (1968) Cellulase activity during the maturation and ripening of tomato fruit. J. Food Sci. 33:588-592.
Huber, D.J., and O'Donoghue, E.M. (1993) Polyuronides in avocado (Persea americana) and tomato (Lycopersicon esculentum) fruits exhibit markedly different patterns of molecular weight downshifts during ripening. Plant Physiol. 102:473-480[Abstract].
Keller, E., and Cosgrove, D.J. (1995) Expansins in growing tomato leaves. Plant J. 8:795-802[Web of Science][Medline].
Kooiman, P. (1960) A method for the determination of amyloid in plant seeds. Recl. Trav. Chim. Pays-Bas Belg. 79:675-678.
Kramer, M., Sanders, R., Bolkan, H., Waters, C., Sheehy, R.E., and Hiatt, W.R. (1992) Postharvest evaluation of transgenic tomatoes with reduced levels of polygalacturonase: Processing, firmness and disease resistance. Postharvest Biol. Technol. 1:241-255[CrossRef].
Laemmli, U.K. (1970) Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680-685[CrossRef][Medline].
Langley, K.R., Martin, A., Stenning, R., Murray, A.J., Hobson, G.E., Schuch, W.W., and Bird, C.R. (1994) Mechanical and optical assessment of the ripening of tomato fruit with reduced polygalacturonase activity. J. Sci. Food Agric. 66:547-554[CrossRef].
Lashbrook, C.C., Gonzalez-Bosch, C., and Bennett, A.B. (1994) Two divergent endo-ß-1,4-glucanase genes exhibit overlapping expression in ripening fruit and abscising flowers. Plant Cell 6:1485-1493[Abstract].
Lashbrook, C.C., Giovannoni, J.J., Hall, B.D., Fischer, R.L., and Bennett, A.B. (1998) Transgenic analysis of tomato endo-ß-1,4-glucanase gene function. Role of cel1 in floral abscission. Plant J. 13:303-310[CrossRef].
Li, Z.-C., Durachko, D.M., and Cosgrove, D.J. (1993) An oat coleoptile wall protein that induces wall extension in vitro and that is antigenically related to a similar protein from cucumber hypocotyls. Planta 191:349-356.
Maclachlan, G., and Brady, C. (1994) Endo-1,4-ß-glucanase, xyloglucanase and xyloglucan endo-transglycosylase activities versus potential substrates in ripening tomatoes. Plant Physiol. 105:965-974[Abstract].
McQueen-Mason, S.J., and Cosgrove, D.J. (1994) Disruption of hydrogen bonding between plant cell wall polymers by proteins that induce wall extension. Proc. Natl. Acad. Sci. USA 91:6574-6578[Abstract/Free Full Text].
McQueen-Mason, S.J., and Cosgrove, D.J. (1995) Expansin mode of action on cell walls. Analysis of wall hydrolysis, stress relaxation, and binding. Plant Physiol. 107:87-100[Abstract].
McQueen-Mason, S.J., Durachko, D.M., and Cosgrove, D.J. (1992) Two endogenous proteins that induce cell wall extension in plants. Plant Cell 4:1425-1433[Abstract/Free Full Text].
McQueen-Mason, S.J., Fry, S.C., Durachko, D.M., and Cosgrove, D.J. (1993) The relationship between xyloglucan endotransglycosylase and in vitro cell wall extension in cucumber hypocotyls. Planta 190:327-331[Web of Science][Medline].
Napoli, C., Lemieux, C., and Jorgensen, R. (1990) Introduction of a chimeric chalcone synthase gene into petunia results in reversible co-suppression of homologous genes in trans.. Plant Cell 2:279-289[Abstract/Free Full Text].
Rose, J.K.C., Lee, H.H., and Bennett, A.B. (1997) Expression of a divergent expansin gene is fruit-specific and ripening-regulated. Proc. Natl. Acad. Sci. USA 94:5955-5960[Abstract/Free Full Text].
Sakurai, N., and Nevins, D.J. (1993) Changes in physical properties and cell wall polysaccharides of tomato (Lycopersicon esculentum) pericarp tissue. Physiol. Plant. 89:681-686[CrossRef].
Seymour, G.B., Lasslett, Y., and Tucker, G.A. (1987) Differential effects of pectolytic enzymes on tomato polyuronides in vivo and in vitro. Phytochemistry 26:3137-3139[CrossRef].
Seymour, G.B., Colquhoun, I.J., DuPont, M.S., Parsley, K.R., and Selvendran, R.R. (1990) Composition and structural features of cell wall polysaccharides from tomato fruits. Phytochemistry 29:725-731[CrossRef].
Shcherban, T.Y., Shi, J., Durachko, D.M., Guiltinan, M.J., McQueen-Mason, S.J., Shieh, M., and Cosgrove, D.J. (1995) Molecular cloning and sequence analysis of expansins—A highly conserved, multigene family of proteins that mediate cell wall extension in plants. Proc. Natl. Acad. Sci. USA 92:9245-9249[Abstract/Free Full Text].
Sheehy, R.E., Kramer, M., and Hiatt, W.R. (1988) Reduction of polygalacturonase activity in tomato fruit by antisense RNA. Proc. Natl. Acad. Sci. USA 85:8805-8809[Abstract/Free Full Text].
Smith, C.J.S., Watson, C.F., Ray, J., Bird, C.R., Morris, P.C., Schuch, W., and Grierson, D. (1988) Antisense RNA inhibition of polygalacturonase gene expression in transgenic tomatoes. Nature 334:724-726[CrossRef].
Smith, D.L., Starrett, D.A., and Gross, K.C. (1998) A gene coding for tomato fruit ß-galactosidase II is expressed during fruit ripening. Plant Physiol. 117:417-423[Abstract/Free Full Text].
Spiers, J., Lee, E., Brady, C.J., Robertson, J., and McGlasson, W.B. (1989) Endopolygalacturonase: Messenger RNA, enzyme and softening in the ripening fruit of a range of tomato genotypes. J. Plant Physiol. 135:576-582.
Stam, M., de Bruin, R., Kenter, S., van der Hoorn, R.A.L., van Blokland, R., Mol, J.N.M., and Kooter, J.M. (1997) Post-transcriptional silencing of chalcone synthase in Petunia by inverted transgene repeats. Plant J. 12:63-82.
Tieman, D.M., and Handa, A.K. (1994) Reduction in pectin methylesterase activity modifies tissue integrity and cation levels in ripening tomato (Lycopersicon esculentum Mill.) fruits. Plant Physiol. 106:429-436[Abstract].
Tieman, D.M., Harriman, R.W., Ramamohan, G., and Handa, A.K. (1992) An antisense pectin methylesterase gene alters pectin chemistry and soluble solids in tomato fruit. Plant Cell 4:667-679[Abstract/Free Full Text].
Tong, C.B.S., and Gross, K.C. (1988) Glycosyl-linkage composition of tomato fruit cell wall hemicellulosic fractions during ripening. Physiol. Plant. 74:365-370[CrossRef].
Tucker, G.A., Robertson, N.G., and Grierson, D. (1980) Changes in polygalacturonase isozymes during the "ripening" of normal and mutant tomato fruit. Eur. J. Biochem. 112:119-124[Web of Science][Medline].
Waterhouse, P.M., Graham, M.W., and Wang, M.-B. (1998) Virus resistance and gene silencing in plants can be induced by simultaneous expression of sense and antisense RNA. Proc. Natl. Acad. Sci. USA 95:13959-13964[Abstract/Free Full Text].
Watson, C.F., Zheng, L., and DellaPenna, D. (1994) Reduction of tomato polygalacturonase ß subunit expression affects pectin solubilization and degradation during fruit ripening. Plant Cell 6:1623-1634[Abstract].
Wu, Y., Sharp, R.E., Durachko, D.M., and Cosgrove, D.J. (1996) Growth maintenance of the maize primary root at low water potentials involves increases in cell-wall extension properties, expansin activity, and wall susceptibility to expansins. Plant Physiol. 111:765-772[Abstract].
Yoder, J.I., Palys, J., Alpert, K., and Lassner, M. (1988) Ac transposition in transgenic tomato plants. Mol. Gen. Genet. 213:291-296.
ABSTRACT
INTRODUCTION
