Antisense Suppression of Deoxyhypusine Synthase ... - Plant Physiology

Antisense Suppression of Deoxyhypusine Synthase in Tomato Delays Fruit Softening and Alters Growth and Development1 Tzann-Wei Wang, Chun-Guang Zhang, Wendy Wu, Linda M. Nowack, Ewa Madey, and John E. Thompson* Department of Biology, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1

The effects of suppressing deoxyhypusine synthase (DHS) have been examined in tomato (Solanum lycopersicum cv UCT5). DHS mediates the first of two sequential enzymatic reactions that activate eukaryotic translation initiation factor-5A (eIF-5A) by converting a conserved Lys to the unusual amino acid, deoxyhypusine. DHS protein levels were suppressed in transgenic plants by expressing the 3#-untranslated region of tomato DHS under regulation of the constitutive cauliflower mosaic virus promoter. Fruit from the transgenic plants ripened normally, but exhibited delayed postharvest softening and senescence that correlated with suppression of DHS protein levels. Northern-blot analysis indicated that all four gene family members of tomato eIF-5A are expressed in fruit, and that three are up-regulated in parallel with enhancement of DHS mRNA as the fruit begin to senesce and soften. Transgenic plants in which DHS was more strongly suppressed were male sterile, did not produce fruit, and had larger, thicker leaves with enhanced levels of chlorophyll. The activity of PSII was 2 to 3 times higher in these transgenic leaves than in corresponding leaves of wild-type plants, and there was also enhanced deposition of starch in the stems. The data collectively indicate that suppression of DHS has pleiotropic effects on growth and development of tomato. This may, in turn, reflect the fact that there is a single DHS gene in tomato and that its cognate protein is involved in the activation of four distinct isoforms of eIF-5A.

Fruit ripening and softening are complex developmental processes resulting in changes in color, flavor, aroma, and texture. In the case of tomato fruit (Solanum lycopersicum cv UCT5), for example, the conversion of chloroplasts to chromoplasts containing lycopene accounts for the progressive color changes during ripening (Harris and Spurr, 1969; Bathgate et al., 1985). Numerous new mRNA species are synthesized during tomato fruit ripening and have been found to encode a range of proteins including enzymes required for ethylene biosynthesis, cell wall degradation, and the accumulation of pigments, sugars, volatiles, and organic acids (Grierson et al., 1985; Gray et al., 1992; Brummell and Harpster, 2001). The seminal role of ethylene in climacteric fruit ripening is compellingly illustrated by the finding that transgenic tomato fruit with suppressed ethylene biosynthesis exhibit delayed ripening (Klee, 1993). Cell wall modifications are a key feature of fruit ripening. They include extensive de-esterification and depolymerization of pectin polymers mediated in part by polygalacturonase, which is strongly up-regulated in ripening tomato fruit (Della Penna et al., 1989; Sitrit and Bennett, 1998). Experiments with transgenic tomato fruit in which polygalacturonase mRNA and

1

This work was supported by the Natural Sciences and Engineering Research Council of Canada. * Corresponding author; e-mail [email protected]; fax 519–746–2543. Article, publication date, and citation information can be found at www.plantphysiol.org/cgi/doi/10.1104/pp.105.060194. 1372

protein levels were suppressed have demonstrated, though, that the softened texture of a ripe fruit does not result directly from polygalacturonase-mediated modifications to the pectin network (Brummell and Harpster, 2001). Down-regulation of polygalacturonase does, however, delay fruit senescence and enhance resistance to postharvest pathogens (Langley et al., 1994). Pectin methylesterase activity also increases in tomato fruit during the early ripening stages (Harriman et al., 1991), but although its suppression in transgenic fruit resulted in reduced pectin depolymerization, there was again no effect on firmness during ripening (Tieman and Handa, 1994). Antisense suppression of pectin methylesterase in tomato does, however, result in a dramatic loss of tissue integrity during senescence of stored fruit (Tieman and Handa, 1994). Softening accompanying ripening proved to be significantly reduced in transgenic tomato fruit with suppressed b-galactosidase, an enzyme that is normally up-regulated during the early stages of ripening and serves to remove pectic galactan side chains (Smith et al., 2002). In addition, expansin proteins appear to play an integral role in fruit softening, probably by disrupting hydrogen bonding between cellulose microfibrils and matrix polysaccharides, resulting in loosening of the cell wall structure (Brummell and Harpster, 2001). In this study, we report that suppression of deoxyhypusine synthase (DHS; EC 2.5.1.46), an enzyme thought to be present in all eukaryotic cells (Jenkins et al., 2001), delays loss of tissue integrity in senescing tomato fruit. DHS catalyzes the first of two sequential reactions that result in the conversion of a conserved

Plant Physiology, July 2005, Vol. 138, pp. 1372–1382, www.plantphysiol.org Ó 2005 American Society of Plant Biologists

DHS Suppression Delays Tomato Fruit Softening

Lys in inactive eukaryotic translation initiation factor5A (eIF-5A) to the unusual amino acid, hypusine (Park et al., 1993, 1997). Hypusinated eIF-5A appears to facilitate protein synthesis by acting as a nucleocytoplasmic shuttle protein, selectively translocating specific subsets of mRNAs from the nucleus to the cytoplasm for translation (Bevec and Hauber, 1997). One of the first indications of its selective involvement in protein synthesis came from yeast (Saccharomyces cerevisiae) studies in which the single yeast gene encoding DHS was mutated. The mutant proved incapable of cell division, but otherwise remained viable and showed only a marginal reduction in protein synthesis (Park et al., 1993, 1997; Kang and Hershey, 1994). The same phenotype was achieved by inactivating both gene family members of yeast eIF-5A (Tome and Gerner, 1997; Park et al., 1998). Inhibition of DHS in mammalian cells induces cell cycle arrest (Hanauske-Abel et al., 1995; Jenkins et al., 2001; Lee et al., 2002). Thus, DHS-mediated activation of eIF-5A appears to be required for division of mammalian cells, although there is also recent evidence that a gene family member of eIF-5A in mammalian cells is involved in apoptosis (Li et al., 2004; Taylor et al., 2004). Full-length cDNA clones encoding DHS and eIF-5A have been isolated from a number of plant species (Chamot and Kuhlemeier, 1992; Ober and Hartmann, 1999; Wang et al., 2001, 2003). Moreover, the recent isolation of DHS from a cDNA expression library prepared from osmotically stressed tomato leaves (Wang et al., 2001) indicates that plant DHS in conjunction with eIF-5A may be involved in facilitating the translation of proteins required for cell death. This contention is further supported by analyses of mRNA abundance indicating that DHS is up-regulated in parallel with eIF-5A in a number of senescing plants (Wang et al., 2001, 2003). In this study, a transgenic approach was used to demonstrate the involvement of DHS in tomato fruit senescence. Antisense suppression of DHS in transgenic plants delayed the onset of fruit senescence and, at high levels of suppression, also resulted in male sterility and alterations in leaf size and function.

RESULTS Antisense 3#-Untranslated Region Tomato DHS Transgenic Plants

The 3#-untranslated region (UTR) of tomato DHS (SlDHS) cDNA (Wang et al., 2001; GenBank accession no. AF296078) was subcloned into pKYLX71 in the antisense orientation. Expression of the transgene was regulated by two consecutive copies of the cauliflower mosaic virus (CaMV) 35S promoter. Southern-blot analysis revealed only one copy of the endogenous DHS gene in HindIII-digested genomic DNA from wild-type plants (Fig. 1). Nine primary transformants expressing 35S::antisense 3#-UTR SlDHS (TP1–9) were Plant Physiol. Vol. 138, 2005

Figure 1. Southern analysis of genomic leaf DNA isolated from wildtype and transgenic tomato plants. The transgenic plants are primary transformants. Lanes correspond to different transgenic lines (1–9) expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter, wild-type plants (WT), and a null line (null). Genomic DNA was digested with HindIII, and the Southern blot was probed with tomato 3#-UTR DHS cDNA. Each lane contained 10 mg of DNA. Transgenic lines denoted by an asterisk did not produce fruit.

obtained, and each of these contained one or two copies of the transgene (Fig. 1). T4 plants for three of the transgenic lines, TP3, TP5, and TP7, were obtained by screening successive generations on kanamycin. A fourth line, TP4, which was one of five transgenic T1 plants (primary transformants) that did not produce fruit (Fig. 1), was propagated vegetatively for further study. A null line, which was kanamycin resistant but had lost the antisense DHS gene, was also analyzed (Fig. 1). In keeping with the fact that CaMV 35S is a constitutive promoter, western-blot analysis confirmed that the presence of the transgene decreased levels of endogenous DHS protein in both leaves and fruit. DHS protein was detectable in mature leaves of TP3, TP5, and TP7 plants, but the level of leaf protein was clearly lower for each of the transgenic lines than for corresponding wild-type plants (Fig. 2A). Of particular note, however, is the finding that DHS was more strongly suppressed in leaves of TP4 plants, which did not produce fruit, than in leaves of the fruit-bearing TP3, TP5, and TP7 plants (Fig. 2, A and B). This suggests that high levels of DHS suppression disrupt fruit set. DHS protein also proved to be dramatically lower in both breaker and red-soft fruit for all three fruitbearing transgenic lines (TP3, TP5, and TP7) than in corresponding fruit from wild-type plants (Fig. 2C). DHS Suppression Delayed Fruit Softening

There were no discernible differences in phenotype during growth and development between wild-type plants and those of the fruit-bearing transgenic lines (TP3, TP5, and TP7). However, in keeping with the fact that DHS is up-regulated in softening tomato fruit from wild-type plants (Wang et al., 2001), fruit from these transgenic lines exhibited delayed postharvest wrinkling in comparison to wild-type control fruit. This phenotype segregated with the antisense transgene and is illustrated in Figure 3 for T4 fruit of the TP3 line. Wild-type and transgenic fruit were harvested at the breaker stage and stored at 19°C in 70% 1373

Wang et al.

Figure 2. Western blots of total protein isolated from leaves and fruit of wild-type and transgenic tomato plants. The blots were probed with polyclonal antibodies raised against tomato DHS recombinant protein; each lane contained 10 mg protein. A, Comparative levels of leaf DHS protein in wild-type plants (WT) and in plants of three transgenic tomato lines, TP3, TP5, and TP7, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. Mature, fully expanded leaves were used for the analyses. B, Comparative levels of leaf DHS protein in wild-type plants (WT) and in plants of the transgenic line, TP4, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. C, Comparative levels of DHS protein in tomato fruit at breaker and red-soft stages of development for wild-type plants (WT) and plants of the transgenic lines, TP3, TP5, and TP7. The fruit were harvested at the breaker stage of development and analyzed immediately after harvest and also at the red-soft stage of development, 21 d postbreaker for wild-type fruit, 45 d postbreaker for TP3 fruit, and 34 d postbreaker for TP5 and TP7 fruit.

to 90% relative humidity. Both types of fruit ripened normally under these postharvest storage conditions, but the control fruit showed visible signs of softening within 14 d and, by 28 d after harvest, were extensively wrinkled and at an advanced stage of deterioration (Fig. 3). By contrast, no visible signs of wrinkling or softening were evident for TP3 fruit stored under identical conditions up to 44 d after harvest (Fig. 3). Quantitative measurements of the delay in fruit softening were obtained using a fruit pressure tester. The pressure required to break through a peeled region of the fruit surface, which was taken to be a measure of firmness, was measured at 3-d intervals after harvest. For wild-type fruit, the pressure decreased from approximately 0.8 kg cm22 to approximately 0.4 kg cm22 within 20 d of harvest and continued to drop, reaching approximately 0.3 kg cm22 by day 40 after harvest (Fig. 4A). By contrast, for TP3 fruit, the pressure remained at approximately 0.8 kg cm22 for 20 d after harvest, and by 40 d had only decreased to approximately 0.6 kg cm22 (Fig. 4A). At pressure readings of 0.6 to 0.8 kg 1374

cm22, the fruit were firm and there was no evidence of skin wrinkling. However, at readings of 0.5 kg cm22 and below, the fruit were soft and wrinkled. Thus, TP3 fruit remained firm well past the point where control fruit had begun to soften and spoil. Postharvest TP5 and TP7 fruit also exhibited delayed softening in comparison to wild-type control fruit, but the delay was less pronounced than that for TP3 fruit (Fig. 4, B and C). For each of TP3, TP5, and TP7, the lines depicting changes in fruit firmness during postharvest storage (Fig. 4) are significantly different from those of wild-type fruit based on a two-way ANOVA (P , 0.0001). As well, the delay in softening was not attributable to differential water loss between control and transgenic fruit. For both TP3 fruit, the transgenic line showing the most pronounced delay in softening (Fig. 4), and wild-type control fruit, the decrease in fresh weight reflecting loss of water between the breaker stage of development when the fruit were harvested and 42 d postharvest was approximately 9% (Fig. 5A). Moreover, the rates of water loss over this period were not significantly different for the two types of fruit (Fig. 5A). Two additional parameters of postharvest development were compared for TP3 and wild-type fruit. Polygalacturonase activity in tomato fruit is augmented during ripening and continues to increase as the fruit become over-ripe and begin to soften (Hobson, 1964; Tucker et al., 1980; Brummell and Harpster, 2001). In this study, polygalacturonase activity proved to be lower in the transgenic fruit than in the wild-type control fruit during postharvest development. At 7 d postbreaker, the activity in wildtype fruit was 0.48 mmol min21 mg21 protein (Fig. 5B), which compares favorably with the value of 0.35 mmol min21 mg21 protein reported previously

Figure 3. Photographs illustrating the effect of suppressed DHS protein abundance on postharvest spoilage of tomato fruit. WT, Fruit from wildtype plants; TP3, fruit from the transgenic line, TP3, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. The fruit were harvested at the breaker stage of development and stored in continuous light at 19°C 6 1°C, 70% to 90% relative humidity. Time after harvest is specified in days. The onset of spoilage is indicated by wrinkling of the fruit skin. Plant Physiol. Vol. 138, 2005


mg21 protein by day 21, at which point the fruit were past the over-ripe stage and visibly senescent, and to 0.01 mmol min21 mg21 protein by day 40, at which point the fruit were extensively senescent (Fig. 5B). By contrast, for the transgenic TP3 fruit, polygalacturonase activity was only 33% of that for wild-type fruit at day 7 postbreaker and 20% of that for wild-type fruit at day 21 postbreaker, and at both these stages of development the transgenic fruit were red-firm (Fig. 5B). Moreover, the activity of polygalacturonase remained

Figure 4. The effect of suppressed DHS protein abundance on changes in tomato fruit firmness during postharvest storage. The fruit were harvested at the breaker stage of development and stored in continuous light at 19°C 6 1°C, 70% to 90% relative humidity. Time after harvest is specified in days. Fruit firmness was measured using a pressure tester and is expressed as the pressure (kg cm22) required to puncture the pulp after peeling away the skin. A, Fruit from wild-type plants (WT) and from plants of the transgenic line, TP3, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. B, Fruit from wild-type plants (WT) and from plants of the transgenic line, TP5, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. C, Fruit from wild-type plants (WT) and from plants of the transgenic line, TP7, expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. The data are expressed as means 6 SE for n 5 3. For A to C, the lines depicting changes in fruit firmness during postharvest storage are significantly different (P , 0.0001) based on a two-way ANOVA (Prism 4 for Windows; GraphPad Software, San Diego).

for this enzyme in tomato fruit at 7 d postbreaker (Smith et al., 1990). At this stage of development, the wild-type fruit were red-firm. Earlier studies have shown that polygalacturonase activity of tomato fruit peaks between 7 and 14 d postbreaker, and that polygalacturonase mRNA levels decline by approximately 75% between 7 and 10 d postbreaker (Smith et al., 1990). In keeping with this, polygalacturonase activity for wild-type fruit declined to 0.24 mmol min21 Plant Physiol. Vol. 138, 2005

Figure 5. Effects of suppressed DHS protein abundance on changes in tomato fruit fresh weight, polygalacturonase activity, and electrolyte leakage during postharvest storage. The fruit were harvested at the breaker stage of development. A, Changes in fruit fresh weight. Values are means 6 SE for n 5 3. The lines depicting changes in fresh weight for wild-type fruit and TP3 fruit are not significantly different (P , 0.05) based on a two-way ANOVA (Prism 4 for Windows; GraphPad Software). B, Polygalacturonase activity in cell wall protein extracts of wild-type (WT) and transgenic (TP3) fruit during postharvest storage. Values are means 6 SE for n 5 3. C, Electrolyte leakage from pericarp discs. Values are means 6 SE for n 5 5. Asterisks denote significant differences (P , 0.0001) between the values for wild-type fruit and TP3 fruit at specific time points based on a two-way ANOVA (Prism 4 for Windows; GraphPad Software). 1375

Wang et al.

essentially unchanged for the transgenic TP3 fruit between days 21 and 40 postbreaker in keeping with the fact that the fruit remained red-firm and did not senesce during this period, whereas polygalacturonase activity for wild-type fruit declined by 96% over the same period (Figs. 5B). This large reduction in polygalacturonase activity for wild-type fruit between days 21 and 40 postbreaker presumably reflects progressive deterioration of the fruit as they continue to senesce (Figs. 3 and 4A). In addition, the onset of electrolyte leakage, which is thought to reflect senescence-induced lipase activity leading to lipid phase changes and membrane leakiness (Eze et al., 1986; Sharom et al., 1994), was delayed in the TP3 fruit (Fig. 5C). For wild-type fruit 20 d after harvest at the breaker stage of development, electrolyte leakage from pericarp discs had increased by 132% of the corresponding background rate for breaker fruit (Fig. 5C). By contrast, leakage from pericarp discs of TP3 fruit 20 d after harvest had increased by only 56% of the background rate for corresponding breaker fruit (Fig. 5C). By day 40 after harvest, electrolyte leakage had increased by 216% of the breaker-stage value for wildtype fruit and by only 143% for TP3 fruit (Fig. 5C). Indeed, the leakage of transgenic TP3 pericarp discs at 40 d postbreaker was comparable to that for wild-type pericarp discs at 20 d postbreaker (Fig. 5C). DHS Suppression Results in Male Sterility and Altered Leaf Morphology and Function

Unlike TP3, TP5, and TP7 plants for which there was no discernible phenotypic difference from wild-type plants during growth and development, TP4 plants, which had higher levels of DHS suppression (Fig. 2, A and B), formed flowers, but did not produce fruit and had to be propagated vegetatively. Pollen from TP4 plants was examined microscopically and found to be malformed, suggesting incomplete or abnormal development and male sterility (Fig. 6). That TP4 flowers were male sterile was confirmed by the finding that cross-pollination with wild-type pollen resulted in the formation of fruit (data not shown). Vegetatively propagated TP4 plants also exhibited other phenotypes that distinguished them from wildtype control plants as well as from TP3, TP5, and TP7 transgenic plants. In particular, TP4 plants had fewer internodes (38% less on average) than wild-type plants (data not shown), and the mature leaves were visually larger with more leaflets than those of wild-type plants (Fig. 7A). Indeed, quantitative measurements of leaflet area revealed that mature and senescing leaflets of TP4 plants were 2.4 and 1.4 times larger, respectively, than those of corresponding wild-type leaflets (Table I). There were also differences between TP4 and wildtype plants in leaflet thickness. Specifically, the young and mature leaflets of TP4 plants were, on average, 1.5 and 1.2 times thicker than those of corresponding wild-type plants (Table I). Leaf chlorophyll was also higher in TP4 plants than in wild-type plants. Levels of 1376

Figure 6. Photograph illustrating differences in pollen morphology between 18-week-old wild-type (WT) and transgenic TP4 plants expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. Bar 5 25 mm.

chlorophyll a/b per square centimeter of leaflet surface were 16% and 42% higher in young and mature leaves, respectively , of TP4 plants than in corresponding leaflets of wild-type plants, and were also higher in the TP4 leaflets of mature and old leaves when expressed relative to fresh weight (Table I). The increased chlorophyll a/b levels in TP4 leaves correlated with higher photosynthetic capacity. Depending on the age of the leaves, PSII activity proved to be 2.3- to 3-fold higher for TP4 leaves than for corresponding wild-type leaves when normalized to leaf surface area, and 1.7- to 2.3-fold higher when normalized to leaf fresh weight (Table I). This is consistent with the higher chlorophyll levels per unit surface area and per unit leaf fresh weight in TP4 plants. Of particular interest, however, is the finding, for young leaves only, that PSII activity normalized to chlorophyll was still higher (1.9-fold) for TP4 plants than for wild-type plants (Table I). This difference was not evident for mature and old leaves (Table I). The enhanced photosynthetic capacity of TP4 plants corresponded with increased starch deposition. Tomato plants store starch in the pith of the stems and, in keeping with higher levels of starch deposition, the main stems of TP4 plants proved to be significantly thicker than those of corresponding wildtype plants. The circumference of the main stems of 18-week-old wild-type and TP4 plants was measured at 20-cm intervals. For the first 20 cm, there was no difference in stem thickness, but in the upper reaches of the plant, 60 to 120 cm from the bottom, the circumference of TP4 stems was, on average, 25% larger than the circumference of wild-type plants (Fig. 7B). Moreover, examination of stem cross-sections stained for starch revealed that the increased thickness Plant Physiol. Vol. 138, 2005


Figure 7. Effects of suppressed DHS protein abundance on tomato plant leaf biomass and stem circumference. A, Photograph illustrating leaf biomass for 18-week-old wild-type plants and transgenic TP4 plants. B, Stem circumference over the first 120 cm of stem height for 18-week-old wild-type plants and transgenic TP4 plants. Values are means 6 SE for n 5 3. C, Stem sections from 18-week-old wild-type and transgenic TP4 plants. The sections were cut 125 cm from the base of the plant and stained with I2 to depict starch. Bar 5 1 cm. TP4 plants contain the 3#-UTR of tomato DHS in the antisense orientation expressed under regulation of the CaMV 35S promoter.

of TP4 stems reflected an increase in size of the pith, which was filled with starch (Fig. 7C). Indeed, quantitation of starch levels over the length of the stem indicated that TP4 plants have 78% more starch than corresponding wild-type plants. Wild-type plants contained 1.58 6 0.14 mg starch cm21 of stem length, whereas transgenic TP4 plants contained 2.81 6 0.07 mg starch cm21 stem length. Temporal and Spatial mRNA Abundance Patterns of DHS and eIF-5A Gene Family Members

There is only one DHS gene (GenBank accession no. AF296077) in tomato, but there are four gene family members for tomato eIF-5A: SleIF-5A1 (accession no. AF296083), SleIF-5A2 (accession no. AF296084), SleIF-5A3 (accession no. AF296085), and SleIF-5A4 (accession no. AF296086; Wang et al., 2001). Northern analyses in which total RNA blots were probed with cDNAs corresponding to the coding regions of DHS Plant Physiol. Vol. 138, 2005

and eIF-5A have indicated that both genes are expressed in senescing cotyledons, leaves, flowers, and fruit of the tomato plant (Wang et al., 2001). This parallel expression is in keeping with the contention that DHS mediates the first of two reactions resulting in post-translational activation of eIF-5A (Park et al., 1993, 1997). However, probing with the coding region of eIF-5A does not distinguish between its various gene family members (Wang et al., 2001). Thus, in an effort to gain an increased understanding of why suppression of DHS in transgenic tomato plants has pleiotropic effects on growth and development, temporal and spatial patterns of mRNA abundance for the individual tomato eIF-5A gene family members were determined and compared to those for DHS. The coding regions of the SleIF-5As are 70% to 80% identical at the nucleotide level (Wang et al., 2001). However, the 3#-UTR regions are only 39% to 49% identical (Table II), and hence it proved possible to distinguish the gene family members in northern blots by probing with the respective 3#-UTRs. DHS is expressed at low levels in breaker and redfirm tomato fruit and is strongly up-regulated at the red-soft stage coincident with the onset of senescence (Fig. 8A; Wang et al., 2001). This is consistent with the finding that suppression of DHS delays fruit softening (Figs. 3 and 4). All of the SleIF-5A gene family members are also expressed in fruit at each of these stages of development (Fig. 8A). However, in keeping with the enhanced abundance of DHS mRNA in senescing fruit, three of the eIF-5A gene family members, SleIF-5A2, SleIF-5A3, and SleIF-5A4, show significant up-regulation at the red-soft stage (Fig. 8A). Similarly, a pronounced up-regulation of DHS in senescing tomato flowers correlates with increased mRNA abundance for two eIF-5A gene family members, SleIF-5A1 and SleIF-5A4 (Fig. 8B) and is consistent with the finding that strong suppression of DHS resulted in deformed pollen (Fig. 6). DHS is also constitutively expressed in tomato leaves, but levels of its cognate mRNA increase coincident with the onset of premature senescence induced by osmotic stress (Fig. 8C; Wang et al., 2001). This again correlates with up-regulation of eIF-5A, in this case a single gene family member, SleIF-5A2 (Fig. 8C). Thus, in the event of both natural senescence, as exemplified by softening tomato fruit and senescing tomato flowers (Fig. 8, A and B), and stress-induced premature senescence, as exemplified by osmotically stressed tomato leaves (Fig. 8C), there is strong up-regulation of DHS accompanied by enhanced mRNA abundance for one or more gene family members of SleIF-5A.

DISCUSSION

It has been proposed that plant DHS, like its mammalian and yeast counterparts, mediates the first of two reactions required for the post-translational activation of eIF-5A, specifically the formation of 1377

Wang et al.

Table I. Morphological and physiological comparisons of wild-type and transgenic TP4 tomato plants TP4 plants were expressing the 3#-UTR of tomato DHS in the antisense orientation under regulation of the CaMV 35S promoter. Cuttings 10 cm in length were rooted in Promix BX soil, and the plants were evaluated at 18 weeks of age. Young, mature, and old denote young, mature, and old leaves, respectively. Values are means 6 SE for n 5 3. NS, No significant difference; Chl, chlorophyll; FW, fresh weight. *, Significant differences at P , 0.05, based on a two-tailed t test (Prism 4 for Windows; GraphPad Software); **, significant differences at P , 0.005, based on a two-tailed t test (Prism 4 for Windows; GraphPad Software); ***, significant differences at P , 0.0005, based on a two-tailed t test (Prism 4 for Windows; GraphPad Software). Young

No. of leaflets per leaf Leaflet area (cm2 leaflet21) Thickness of leaf (nm) Chl a/b/leaf area (mg cm22) Chl a/b/FW (mg g21 FW) PSII/leaf area (mg O2 h21 cm22) PSII/FW (mg O2 h21 g21 FW) PSII/chl (mg O2 h21 mg21 chl a/b)

Wild type TP4 Wild type TP4 Wild type TP4 Wild type TP4 Wild type TP4 Wild type TP4 Wild type TP4 Wild type TP4

8.7 11.3 12.8 11.3 2.55 3.86 46.6 54.2 1.28 1.17 7.5 16.9 219 378 176 336

6 0.8 6 0.4 6 0.7 6 0.5 6 0.06 6 0.14 6 1.7 6 1.9 6 0.08 6 0.06 6 1.6 6 2.3 6 45.6 6 50.9 6 47.6 6 50.4

deoxyhypusine on the inactive eIF-5A protein (Wang et al., 2003). Evidence supporting this includes the fact that recombinant tomato and tobacco (Nicotiana tabacum) DHS have been shown capable of catalyzing the formation of a deoxyhypusine residue in Arabidopsis (Arabidopsis thaliana) and tobacco eIF-5A substrates, respectively (Ober and Hartmann, 1999; Wang et al., 2001). These observations indicate not only that plant DHS exhibits the same catalytic property as its animal and yeast counterparts (Schwelberger et al., 1993; Park et al., 1997), but also that recombinant plant eIF-5A is capable of being deoxyhypusine modified. In addition, changes in DHS mRNA abundance during plant growth and development are accompanied by corresponding changes in eIF-5A mRNA levels. In tomato, for example, there is strong up-regulation of DHS accompanying flower senescence, softening of tomato fruit, and premature leaf senescence induced by environmental stress, and in each case the enhanced abundance of DHS mRNA is accompanied by upregulation of eIF-5A (Wang et al., 2001). Wang et al. (2001) reported that northern blots probed with either full-length SlDHS cDNA or 3#-UTR SlDHS cDNA depict the same spatial and temporal patterns of mRNA abundance during growth and development of tomato. This could mean that there is only one gene encoding DHS in tomato and is consistent with Southern-blot data obtained in this study. These earlier findings are also consistent with the fact that transgenic plants constitutively expressing antisense 3#-UTR cDNA of the single known SlDHS gene exhibited strong phenotypes. Yet there are four gene family members of SleIF-5A. This suggests that the protein encoded by the single SlDHS gene is involved in the post-translational activation of all of the SleIF-5A isoforms, and that the phenotypes observed in the 1378

t Test

* NS *** * NS * NS *

Mature

11.0 15.8 22.8 54.6 3.69 4.41 60.5 86.0 1.66 2.23 8.2 18.1 238 405 150 187

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

0.8 1.1 1.1 1.4 0.08 0.18 4.62 1.4 0.09 0.14 0.4 2.0 12.7 44.5 11.5 33.4

t Test

Old

**

8.8 10.0 27.1 38.2 4.92 4.98 50.0 63.2 1.22 2.26 5.4 15.7 156 351 175 161

*** * ** * ** * NS

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

t Test

0.6 0.7 0.8 0.9 0.04 0.04 5.4 5.1 0.14 0.12 0.6 0.9 17.3 19.1 23.9 15.9

NS *** NS NS ** *** *** NS

DHS-suppressed transgenic plants are not necessarily a reflection of functional impairment of a single isoform of eIF-5A. Indeed, although up-regulation of DHS in the various tissues of developing tomato plants appears to be invariably accompanied by a corresponding enhancement of eIF-5A mRNA abundance, it is apparent from this study that there is no clear pattern of up-regulation for the individual gene family members of eIF-5A. Rather, the expression patterns suggest that there may be some functional redundancy among SleIF-5A gene family members. For example, in senescing tomato fruit, three gene family members, SleIF-5A2, SleIF-5A3, and SleIF-5A4, show enhanced transcript abundance and, in senescing flowers, SleIF-5A1 and SleIF-5A4 are both up-regulated. Although more definitive testing would be needed to demonstrate that the gene family members of SleIF-5A have even a degree of overlapping function during senescence, it is noteworthy that both gene family members of eIF-5A in yeast regulate cell division (Schwelberger et al., 1993; Park et al., 1997). For three of the DHS-suppressed transgenic lines, TP3, TP5, and TP7, there was a significant delay in the onset of fruit senescence during postharvest storage. This was apparent both from the appearance of the

Table II. 3#-UTR nucleotide sequence identities (%) for tomato eIF-5A gene family members Corresponding GenBank numbers are indicated.

SleIF-5A2 (AF296084) SleIF-5A3 (AF296085) SleIF-5A4 (AF296086)

SleIF-5A1 (AF296083)



43.2 46.4 45.4

– 39.3 49.3

– – 44.2

Plant Physiol. Vol. 138, 2005


Figure 8. Northern-blot analysis of total RNA isolated from tissues and organs of tomato plants. A, Fruit. BK, Breaker; RF, red-firm; RS, red-soft. The fruit were harvested at the breaker stage; the red-firm fruit were analyzed at 7 d postbreaker, and the red-soft fruit at 21 d postbreaker. B, Flowers. FB, Flower buds; SF, fully open and senescing flowers. C, Leaves of 8-week-old tomato seedlings. WC, Leaves from 6-h water-treated derooted tomato seedlings; S, leaves from 6-h sorbitol-treated derooted tomato seedlings. The blots were probed with 3#-UTR cDNAs of the tomato eIF-5A gene family members, SleIF-5A1, SleIF-5A2, SleIF-5A3, and SleIF-5A4, and with full-length cDNA of tomato DHS. ND, Not detectable. Each lane contained 10 mg of RNA. The corresponding ethidium bromide-stained agarose gels of fractionated RNA are illustrated.

fruit, specifically delayed wrinkling and spoilage in comparison with wild-type fruit, and from quantitative measurements showing delayed pericarp softening, and indicates that DHS plays a role in the execution of fruit senescence. Moreover, for each of the lines, the delay in fruit senescence correlated with reduced abundance of DHS protein in the fruit tissue. DHS-suppressed fruit also exhibited a delay in the onset of electrolyte leakage associated with fruit senescence. Electrolyte leakage is a well-established trait of Plant Physiol. Vol. 138, 2005

senescing tissues and is thought to reflect lipid phase changes attributable to enhanced lipase activity and ensuing leakiness of membrane bilayers (Eze et al., 1986; Sharom et al., 1994). Electrolyte leakage for transgenic fruit at 40 d postbreaker was the same as that for wild-type fruit at 20 d postbreaker, confirming a very significant delay in senescence and spoilage of the transgenic fruit. In addition, the transgenic fruit exhibited lower levels of polygalacturonase activity than wild-type control fruit during postharvest development. This enzyme is known to be associated with fruit ripening as well as softening of over-ripe fruit (Hobson, 1964; Tucker et al., 1980; Brummell and Harpster, 2001). Indeed, the polygalacturonase activity of transgenic fruit was only 33% of that for corresponding wild-type fruit at 7 d postbreaker when both types of fruit were still red-firm, and 20% of the activity for wild-type fruit at 21 d postbreaker when the transgenic fruit were still red-firm and the wildtype fruit were soft and extensively senescent. This finding that reduced polygalacturonase activity correlates with enhanced structural integrity of tomato fruit during postharvest storage has been noted previously in studies of transgenic fruit with suppressed polygalacturonase activity (Kramer et al., 1992; Langley et al., 1994; Brummell and Harpster, 2001). Of particular interest is the finding that suppression of DHS had no effect on fruit ripening. Transgenic and wild-type fruit harvested at the breaker stage of development ripened normally and in parallel, and were of comparable texture and color at the fully ripe stage. Thus, the delay in senescence-associated softening of the fully ripe transgenic fruit with suppressed DHS appears to be mechanistically distinguishable from the textural change incurred during tomato fruit ripening, a change that is at least partially mediated by up-regulation of b-galactosidase (Smith et al., 2002). That these temporal phases of fruit softening are executed by different mechanisms is also consistent with the finding that antisense suppression of pectin methylesterase in tomato fruit accelerates the senescence-associated loss of tissue integrity during senescence, but has no effect on fruit firmness during ripening (Tieman and Handa, 1994). Apart from exhibiting delayed fruit senescence, transgenic lines TP3, TP5, and TP7 were indistinguishable from wild-type plants during growth and development. However, an additional DHS-suppressed transgenic line, TP4, exhibited deformed pollen and proved incapable of forming fruit unless crosspollinated with wild-type pollen. Moreover, westernblot analyses of leaf tissue indicated that DHS is more strongly suppressed in TP4 plants than in TP3, TP5, and TP7 plants. Pollen formation is contingent upon a series of highly regulated developmental changes in the flower, some of which involve tissue senescence. DHS is strongly up-regulated in senescing flowers, and the finding that its suppression in TP4 plants disrupts pollen formation is consistent with its apparent involvement in tissue senescence. 1379

Wang et al.

TP4 plants also had larger and thicker leaves with more chlorophyll per unit surface area than wild-type plants, and the higher levels of chlorophyll correlated with enhanced photosynthetic capability. Moreover, this increase in leaf size and thickness and the accompanying increase in photosystem activity were not apparent in TP3, TP5, and TP7 plants, which had lower levels of DHS suppression. The underlying basis for enhanced leaf size in the DHS-suppressed TP4 plants is not clearly apparent, but it may reflect reduced severity of senescence induction and the associated cessation of growth in response to repeated episodes of sublethal stress. This would result in growth that is more continuous over time (Thompson et al., 2004). It is important to note, as well, that inasmuch as TP4 is a primary transformant, tissue culture effects cannot be precluded. Suppression of DHS has been shown previously to delay the onset of leaf senescence in transgenic Arabidopsis plants (Wang et al., 2003). The results of this study, indicating that suppression of DHS in tomato delays the onset of fruit senescence, lend further credence to the view that DHS plays a central role in the regulation of senescence. DHS is thought to be present in all eukaryotic cells and, to date, its role in mammalian cells and yeast appears to be restricted to an involvement in hypusination of eIF-5A. Moreover, there are reports that distinct isoforms of hypusinated mammalian eIF-5A play a role in cell proliferation and apoptosis (Lee et al., 2002; Li et al., 2004; Taylor et al., 2004). This study does not preclude a role for DHS in cell division in plants, but it does indicate that plant DHS is involved in the regulation of senescence, a highly regulated process of programmed cell death analogous to apoptosis.

MATERIALS AND METHODS Suppression of DHS in Transgenic Plants Suppression of endogenous tomato (Solanum lycopersicum cv UCT5) DHS (SlDHS) protein abundance was achieved by constitutively expressing the 3#-UTR of SlDHS cDNA in the antisense orientation in transgenic plants. The 3#-UTR of SlDHS cDNA, which has been cloned previously (Wang et al., 2001), was subcloned into the binary vector, pKYLX71 (Schardl et al., 1987) in the antisense orientation under the regulation of two copies of the CaMV 35S promoter. The pKYLX71 vector contains a tetracycline resistance gene in the bacterial replication region and a kanamycin resistance gene (NPTII) in the T-DNA region. Binary vector bearing the antisense DHS construct was introduced into Agrobacterium tumefaciens LBA 4404 by electroporation. Liquid culture (10 mL 2xYT containing 50 mg tetracycline and 250 mg rifampicin) from a single colony of transformed A. tumefaciens was centrifuged at 2,500g for 5 min, and the sediment was resuspended in liquid Murashige and Skoog (MS) medium (Sigma-Aldrich, Oakville, Ontario, Canada) at 2.5 3 108 bacteria mL21 (OD600 5 0.4) to form an agrosuspension. Tomato cotyledon explants were transformed with agrosuspension essentially as described by Earle and Frary (1996). For this purpose, seeds were surface sterilized in 1% sodium hypochlorite-0.1% Tween 80 for 10 min, rinsed three times in sterile water, and germinated in a Magenta box on half-strength MS media containing 0.7% agar and 1.5% Suc at 21°C under 60 to 100 mmol m22 s21 white fluorescent light with 16-h-light/8-h-dark cycles. Cotyledons were excised from 8-d-old seedlings and placed on wet sterile paper towels. Each cotyledon was cut into three pieces. The explants were precultured on MS media containing zeatin (1 mg L21) and naphthylacetic acid (0.1 mg L21)

1380

with 0.7% agar for 1 d, cocultivated with Agrobacteria for 3 d, and then transferred to the selective regeneration medium (MS media containing zeatin [2 mg L21], kanamycin [50 mg L21], and carbenicillin [200 mg L21]). Callus was transferred to new media every 3 weeks until shoots approximately 2 cm in size were obtained. The shoots were then transferred to Magenta boxes containing rooting media (half-strength MS media, naphthylacetic acid [0.1 mg L21], kanamycin [50 mg L21], and carbenicillin [200 mg L21]). Roots formed within 1 week and the plantlets were transferred to soil 1 week later and grown to maturity in a greenhouse. Segregating populations of seed were screened on kanamycin to obtain transgenic lines. Transgenic and wild-type plants were grown in Promix BX soil in 14-inch pots in a greenhouse. Some transgenic T1 plants produced flowers, but no fruit. These lines were maintained as rooted cuttings under greenhouse conditions for further analysis. Cuttings (10 cm in length) from 18-week-old plants were rooted in Promix BX soil in 8-inch pots and, after 4 weeks, transferred to 14-inch pots and allowed to grow to maturity.

Molecular Analyses Leaf genomic DNA for Southern-blot analysis was isolated according to Wang et al. (2001) and digested with HindIII. The digested products (10 mg DNA) were fractionated on a 1.0% agarose gel, immobilized on a nylon membrane, and hybridized with 3#-UTR SlDHS cDNA as described in Wang and Arteca (1995). Western-blot analysis and protein extraction and measurement were carried out as described previously (Wang et al., 2001). For western-blot analysis of fruit, total fruit protein was subjected to ammonium sulfate precipitation as described in Park and Wolff (1988) prior to fractionation in order to eliminate pectin contaminants. The 25% to 40% ammonium sulfate fraction, which contains the DHS protein (Park and Wolff, 1988), was used for western-blot analysis. For northern-blot analysis, total RNA (10 mg) was fractionated on 1.2% denatured formaldehyde agarose gels and immobilized on Hybond-N1 nylon membrane (Amersham-Pharmacia, Baie d’Urfe, Canada). Hybridization conditions were as described in Wang and Arteca (1995). cDNAs corresponding to the 3#-UTRs of the four tomato eIF-5A gene family members and full-length cDNA corresponding to tomato DHS were labeled with aP32-dATP using a random primer kit (Roche, Mannheim, Germany).

Postharvest Fruit Storage, Firmness, and Electrolyte Leakage Tomato fruit were harvested at the breaker stage of development and stored in continuous light at 19°C and 70% to 90% relative humidity until they ripened and began to senesce. Firmness of the fruit during this period of postharvest storage was measured at 3-d intervals using two different fruit pressure testers (FT02 and FT011; Wagner Instruments, Greenwich, CT). FT02 was used for measuring pressures up to 1 kg cm22, and FT011 was used for measuring pressures between 1 and 5 kg cm22. The convex tip for both testers is five-sixteenths of an inch in diameter. Fruits were peeled at the location of measurement due to the resistance of the skin, and firmness was taken to be the amount of pressure required to puncture the surface of the exposed fruit pulp. Electrolyte leakage of pericarp discs from fruit at selected stages of postharvest development was measured at room temperature over a 3-h period using a conductance meter (model 1056; Amber Science, Eugene, OR) as described previously (Sharom et al., 1994).

Morphological Measurements and Chemical Assays Leaf age was designated according to position on the mature plant. Leaves at the top of the plant, 10 to 11 cm in length, approximately 100 cm from the base of the stem, and light green in color were classified as young leaves. Those in the middle region of the plant, 20 to 100 cm from the base of the stem, 18 to 24 cm in length, and dark green in color were designated as mature leaves. Those at the lower reaches of the plant, within the first 20 cm from the base of the stem, 20 to 27 cm in length, and beginning to turn yellow were designated as senescing leaves. Leaf area was measured with a portable area meter (model LI-300A; LI-COR, Lincoln, NE), and leaf thickness was measured using calipers. Stem circumference was measured at 10-cm intervals by wrapping a piece of string



around the stem to accurately replicate its shape and measuring the length of string corresponding to the circumference. Levels of chlorophyll a/b in leaf discs (0.7 cm diameter) were measured as described by Porra et al. (1989). PSII activity was determined according to Wang et al. (2003). Starch content in the pith of stems from 18-week-old mature plants was visualized by examining 40-mm-thick stem sections that had been stained with Lugol’s solution for approximately 2 min, rinsed in deionized water, mounted on glass slides, and overlaid with a coverslip. The sections were photographed on a light box with a digital camera. Levels of stem starch were quantified according to Lustinec et al. (1983).

Measurements of Polygalacturonase Activity Protein extracts for measurements of polygalacturonase were obtained as described by Biggs and Handa (1988) and Tucker and Grierson (1982). Polygalacturonase activity was measured according to Gross (1982) in the absence of added Ca21. Under these conditions, endo-polygalacturonase activity as distinct from exo-polygalacturonase activity is measured (Burmeister and Harman, 1998). Protein was quantified using Bradford reagent (Bradford, 1976).

ACKNOWLEDGMENT We are grateful to Ms. Lynn Hoyles for expert technical assistance. Received January 25, 2005; revised March 28, 2005; accepted April 12, 2005; published June 10, 2005.

LITERATURE CITED Bathgate B, Purton ME, Grierson D, Goodenough PW (1985) Plastid changes during the conversion of chloroplasts to chromoplasts in ripening tomatoes. Planta 165: 197–204 Bevec D, Hauber J (1997) Eukaryotic initiation factor 5A activity and HIV-1 Rev function. Biol Signals 6: 124–133 Biggs MS, Handa AK (1988) Temporal regulation of polygalacturonase gene expression in fruits of normal, mutant, and heterozygous tomato genotypes. Plant Physiol 89: 117–125 Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248–254 Brummell DA, Harpster MH (2001) Cell wall metabolism in fruit softening and quality and its manipulation in transgenic plants. Plant Mol Biol 47: 311–340 Burmeister DM, Harman JE (1998) Effect of fruit maturity on the success of controlled atmosphere storage of Fantasia nectarines. Acta Hortic 464: 363–368 Chamot D, Kuhlemeier C (1992) Differential expression of genes encoding the hypusine-containing translation initiation factor, eIF-5A, in tobacco. Nucleic Acids Res 20: 665–669 Della Penna D, Lincoln JE, Fischer RL, Bennett AB (1989) Transcriptional analysis of polygalacturonase and other ripening associated genes in Rutgers, rin, nor, and Nr tomato fruit. Plant Physiol 90: 1372–1377 Earle ED, Frary A (1996) An examination of factors affecting the efficiency of Agrobacterium-mediated transformation of tomato. Plant Cell Rep 16: 235–240 Eze JMO, Mayak S, Thompson JE, Dumbroff EB (1986) Senescence in cut carnation flowers—temporal and physiological relationships among water status, ethylene, abscisic-acid and membrane-permeability. Physiol Plant 68: 323–328 Gray J, Picton S, Shabbeer J, Schuch W, Grierson D (1992) Molecular biology of fruit ripening and its manipulation with antisense genes. Plant Mol Biol 19: 69–87 Grierson D, Slater A, Maunders M, Crookes P, Tucker GA, Schuch W, Edwards K (1985) Regulation of the expression of tomato fruit ripening genes: the involvement of ethylene. In JA Roberts, GA Tucker, eds, Ethylene and Plant Development. Robert Hartnoll Ltd., Bodmin, Cornwall, UK, pp 147–161 Gross KC (1982) A rapid and sensitive spectrophotometric method for


assaying polygalacturonase using 2-cyanoacetamide. HortScience 17: 933–934 Hanauske-Abel HM, Slowinska B, Zagulska S, Wilson RC, Staiano-Coico L, Hanauske AR, McCaffrey T, Szabo P (1995) Detection of a sub-set of polysomal mRNAs associated with modulation of hypusine formation at the G1–S boundary. Proposal of a role for eIF-5A in onset of DNA replication. FEBS Lett 366: 92–98 Harriman RW, Tieman DM, Handa AK (1991) Molecular cloning tomato pectin methylesterase gene and its expression in Rutgers, ripening inhibitor, nonripening, and Never Ripe tomato fruits. Plant Physiol 97: 80–87 Harris WM, Spurr AR (1969) Chromoplasts of tomato fruits. II. The red tomato. Am J Bot 56: 380–389 Hobson GE (1964) Polygalacturonase in normal and abnormal tomato fruit. Biochem J 92: 324–332 Jenkins ZA, Haag PG, Johansson HE (2001) Human eIF-5A2 on chromosome 3q25-q27 is a phylogenetically conserved vertebrate variant of eukaryotic translation initiation factor 5A with tissue-specific expression. Genomics 71: 101–109 Kang HA, Hershey JWB (1994) Effect of initiation factor eIF-5A depletion on protein synthesis and proliferation of Saccharomyces cerevisiae. J Biol Chem 269: 3934–3940 Klee HJ (1993) Ripening physiology of fruit from transgenic tomato (Lycopersicon esculentum) plants with reduced ethylene synthesis. Plant Physiol 102: 911–916 Kramer M, Sanders R, Bolkan H, Waters C, Sheehy RE, Hiatt WR (1992) Postharvest evaluation of transgenic tomatoes with reduced levels of polygalacturonase; processing, firmness and disease resistance. Postharvest Biol Technol 1: 241–255 Langley KR, Martin A, Stenning R, Murray AJ, Hobson GE, Schuch WW, Bird CR (1994) Mechanical and optical assessment of the ripening of tomato fruit with reduced polygalacturonase activity. J Sci Food Agric 66: 547–554 Lee Y, Kim HK, Park HE, Park MH, Joe YA (2002) Effect of N-1-guanyl-1,7diaminoheptane, an inhibitor of deoxyhypusine synthase, on endothelial cell growth, differentiation and apoptosis. Mol Cell Biochem 237: 69–76 Li AL, Li HY, Jin BF, Ye QN, Zhou T, Yu XD, Pan X, Man JH, He K, Yu M, et al (2004) A novel eIF5A complex functions as a regulator of p53 and p53-dependent apoptosis. J Biol Chem 279: 49251–49258 Lustinec J, Hadacova V, Kaminek M, Prochazka Z (1983) Quantitative determination of starch, amylose, and amylopectin in plant tissues using glass fiber paper. Anal Biochem 132: 265–271 Ober D, Hartmann T (1999) Deoxyhypusine synthase from tobacco. J Biol Chem 274: 32040–32047 Park MH, Joe YA, Kang KR (1998) Deoxyhypusine synthase activity is essential for cell viability in the yeast Saccharomyces cerevisiae. J Biol Chem 273: 1677–1683 Park MH, Lee YB, Joe YA (1997) Hypusine is essential for eukaryotic cell proliferation. Biol Signals 6: 115–123 Park MH, Wolff EC (1988) Cell-free synthesis of deoxyhypusine. Separation of protein substrate and enzyme and identification of 1,3-diaminopropane as a product of spermidine cleavage. J Biol Chem 263: 15264–15269 Park MH, Wolff EC, Folk JE (1993) Hypusine: its post-translational formation in eukaryotic initiation factor 5A and its potential role in cellular regulation. Biofactors 4: 95–104 Porra RJ, Thompson WA, Kriedemann PE (1989) Determination of accurate extinction coefficients and simultaneous equations for assaying chlorophylls a and b extracted with four different solvents: verification of the concentration of chlorophyll standards by atomic absorption spectroscopy. Biochim Biophys Acta 975: 384–394 Schardl CL, Byrd AD, Benzion G, Altschuler MA, Hildebrand DF, Hunt AG (1987) Design and construction of a versatile system for the expression of foreign genes in plants. Gene 61: 1–11 Schwelberger HG, Kang HA, Hershey JWB (1993) Translation initiation factor eIF-5A expressed from either of two yeast genes or from human cDNA; functional identity under aerobic and anaerobic condition. J Biol Chem 268: 14018–14025 Sharom M, Willemot C, Thompson JE (1994) Chilling injury induces lipid phase changes in membranes of tomato fruit. Plant Physiol 105: 305–308 Sitrit Y, Bennett AB (1998) Regulation of tomato fruit polygalacturonase

1381

Wang et al.

mRNA accumulation by ethylene: a reexamination. Plant Physiol 116: 1145–1150 Smith CJS, Watson CF, Morris PC, Bird CR, Seymour GB, Gray JE, Arnold C, Tucker GA, Schuch W, Harding S, et al (1990) Inheritance and effect on ripening of antisense polygalacturonase genes in transgenic tomatoes. Plant Mol Biol 14: 369–379 Smith DL, Abbott JA, Gross KC (2002) Down-regulation of tomato b-galactosidase 4 results in decreased fruit softening. Plant Physiol 129: 1755–1762 Taylor CA, Senchyna M, Flanagan J, Joyce EM, Cliche DO, Boone AN, Culp-Stewart S, Thompson JE (2004) Role of eIF5A in TNFalpha-mediated apoptosis of lamina cribrosa cells. Invest Ophthalmol Vis Sci 45: 3568–3576 Thompson JE, Hopkins MT, Taylor C, Wang T-W (2004) Regulation of senescence by eukaryotic translation initiation factor 5A: implications for plant growth and development. Trends Plant Sci 9: 174–179 Tieman DM, Handa AK (1994) Reduction in pectin methylesterase activity modifies tissue integrity and cation levels in ripening tomato (Lycopersicon esculentum Mill.) plants. Plant Physiol 106: 429–436

1382

Tome ME, Gerner EW (1997) Cellular eukaryotic initiation factor 5A content as a mediator of polyamine effects of growth and apoptosis. Biol Signals 6: 150–156 Tucker GA, Grierson D (1982) Synthesis of polygalacturonase during tomato fruit ripening. Planta 155: 64–67 Tucker GA, Robertson NG, Grierson D (1980) Changes in polygalacturonase isoenzymes during the ripening of normal and mutant tomato fruit. Eur J Biochem 115: 87–90 Wang T-W, Arteca RN (1995) Identification and characterization of cDNAs encoding ethylene biosynthetic enzymes from Pelargonium 3 hortorum cv Snow Mass leaves. Plant Physiol 109: 627–636 Wang TW, Lu L, Wang D, Thompson JE (2001) Isolation and characterization of senescence-induced cDNAs encoding deoxyhypusine synthase and eucaryotic translation initiation factor-5A from tomato. J Biol Chem 276: 17541–17549 Wang T-W, Lu L, Zhang C-G, Taylor C, Thompson JE (2003) Pleiotropic effects of suppressing deoxyhypusine synthase expression in Arabidopsis thaliana. Plant Mol Biol 52: 1223–1235
