TY - JOUR
AU - Eason, Jocelyn R.
AB - Abstract Harvest‐induced senescence of broccoli results in tissue wilting and sepal chlorosis. As senescence progresses, chlorophyll and protein levels in floret tissues decline and endo‐protease activity (measured with azo‐casein) increases. Protease activity increased from 24 h after harvest for tissues held in air at 20 °C. Activity was lower in floret tissues from branchlets that had been held in solutions of sucrose (2% w/v) or under high carbon dioxide, low oxygen (10% CO2, 5% O2) conditions. Four protease‐active protein bands were identified in senescing floret tissue by zymography, and the use of chemical inhibitors of protease action suggests that some 44% of protease activity in senescing floret tissue 72 h after harvest is due to the action of cysteine and serine proteases. Four putative cysteine protease cDNAs have been isolated from broccoli floret tissue (BoCP1, BoCP2, BoCP3, BoCP4). The cDNAs are most similar (73–89% at the amino acid level) to dehydration‐responsive cysteine proteases previously isolated from Arabidopsisthaliana (RD19, RD21). The mRNAs encoded by the broccoli cDNAs are expressed in floret tissue during harvest‐induced senescence with mRNA accumulating within 6 h of harvest for BoCP1, 12 h of harvest for BoCP4 and within 24 h of harvest for BoCP2 and BoCP3. Induction of the cDNAs is differentially delayed when broccoli branchlets are held in solutions of water or sucrose. In addition, the expression of BoCP1 and BoCP3 is inhibited in tissue held in atmospheres of high carbon dioxide/low oxygen (10% CO2, 5% O2). The putative cysteine protease mRNAs are expressed before measurable increases in endo‐protease activity, loss of protein, chlorophyll or tissue chlorosis. Key words: Brassicaoleracea, broccoli, controlled atmosphere, post‐harvest, senescence, wilting. Received 24 September 2002; Accepted 26 November 2002 Introduction Proteolysis, the degradation of proteins, occurs via multiple pathways and is of fundamental importance for the normal development, homeostasis and final death of a plant cell (reviewed in Vierstra, 1996; Beers et al., 2000; Callis and Vierstra, 2000). The biochemical degradation of proteins through hydrolysis of peptide bonds is caused by the action of proteolytic enzymes or proteases. Protease activity within the cellular compartments (most of which have been shown to contain proteolytic activity, Beers et al., 2000) may be regulated at various levels by transcriptional and translational regulation, by post‐translational processing, and through the action of specific protease inhibitor proteins (Solomon et al., 1999). In plants, the genetically programmed death of a whole organ, for example, a leaf or flower, is designated senescence. During senescence, cellular organelles are dismantled and macromolecules are degraded releasing nutrients for remobilization to other rapidly growing tissues in the plant. Protease action in tissues undergoing senescence provides the plant with transportable amino acids (e.g. asparagine). There is, however, growing evidence that specific proteases may also act as mediators of signal transduction or effectors of programmed cell death during plant senescence (Beers et al., 2000). Through the invitro use of specific protease inhibitors, researchers have attributed protease activity associated with flower senescence to cysteine proteases (Stephenson and Rubinstein, 1998; Eason et al., 2002). In addition, the induction of cysteine protease mRNAs has been associated with senescence of flowers (Jones et al., 1995; Guerrero et al., 1998; Eason et al., 2002; Wagstaff et al., 2002), leaves (Lohman et al., 1994; John et al., 1997; Buchanan‐Wollaston and Ainsworth, 1997), roots (Chevalier et al., 1995), and pea nodules (Kardailsky and Brewin, 1996). Cysteine protease genes are expressed in plant tissues during times of stress, including drought (Koizumi et al., 1993; Williams et al., 1994), salt stress (Jones and Mullet, 1995), cold temperature (Schaffer and Fischer, 1988), wounding (Linthorst et al., 1993), ethylene treatment (Cervantes et al., 1994; Jones et al., 1995), and following glucose starvation (Chevalier et al., 1995). Such stresses may induce senescence. A role for protease inhibitors has been proposed as modulators of programmed cell death in plants (Solomon et al., 1999); the suggestion is that cell death may be regulated by the opposing activities of protease versus protease inhibitor enzymes. The current research was undertaken to characterize the expression of cysteine proteases during harvest‐induced broccoli senescence. With the exception of protease analysis, the physiological and biochemical changes that are associated with senescence in broccoli are well characterized. Before harvest, broccoli is comprised of hundreds of rapidly developing florets at various stages of bud development, attached to a thick fleshy stem. The ‘vegetable’ is harvested when the florets are immature (green), and their deterioration after harvest (senescence) is characterized by a loss of turgor and wilting, chlorosis (particularly the floret sepals), and the production of off‐flavours and smells. Cellular respiration declines rapidly after harvest, as do the levels of protein, soluble and storage carbohydrates, citric acid, and malic acid (all within 24 h of harvest; King and Morris, 1994a). A decline in chlorophyll levels is evident after 48 h of harvest (Clarke et al., 1994), and the ammonia level and free amino acid content increases after 72 and 96 h, respectively (King and Morris, 1994a). The activity of lipoxygenase, the enzyme that generates fatty acid hydroperoxides, increases markedly after 72 h (Page et al., 2001). Senescence‐associated gene expression in broccoli is induced within 2 h of harvest (Pogson et al., 1995; Downs and Somerfield, 1997). A recent molecular analysis has identified a number of novel transcripts showing enhanced expression in broccoli florets and leaves during senescence (Page et al., 2001). In particular, the induction and differential expression of four putative cysteine protease mRNAs during broccoli floret senescence (Page et al., 2001). In order to understand the impact of cysteine protease enzymes on harvest‐induced senescence of immature floral tissues, the water loss from harvested broccoli heads has been evaluated, endo‐protease activity assessed and the expression patterns determined of newly isolated putative dehydration‐responsive cysteine protease genes after harvest and during treatments that slow or inhibit the process of senescence (water, sucrose, controlled atmosphere). Materials and methods Plant material Broccoli (Brassicaoleracea L. var. Italica) was harvested in the cool of the morning from a commercially grown crop. Zero hour floret samples were frozen in liquid nitrogen in the field. Harvested broccoli heads were placed on ice and returned to the laboratory within 1 h of harvest. The heads were dissected, and branchlets of uniform size and development were excised from the second layer above the head base, as previously described (King and Morris, 1994b). Water loss from broccoli tissue after harvest Water loss at 20 °C was monitored by repeated weighing after coating various tissue parts with petroleum jelly or after feeding sucrose (70 mm) or abscisic acid (0–2 mM, 22 h) to the cut end of branchlets. Proton magnetic resonance imaging (MRI) was carried out in a wide‐bore 4.7 Tesla magnet with custom‐designed hardware and software. Plant material was immobilized within perforated perspex cylinders inside 40 mm birdcage coils. A small external fan passed air down the bore of the magnet; exit air velocity was 0.5 m s–1. At intervals, Hahn spin‐echo images of the imaging plane were acquired, using repetition times (Tr) of 100–3200 ms and echo times (Te) of 25–160 ms. Each image of a 20–35 mm diameter field of view was acquired as a 256×256 data array; slice thickness was 3 mm. Calculated images of relaxation parameters (T1, T2) and pseudo‐proton spin density were determined by computation as previously described (Clark et al., 1998). Post‐harvest treatments to delay senescence In the first experiment, branchlets were held dry, in water or in sucrose (2% w/v, 58 mM). The onset of chlorosis was determined by measuring sepal colour (using a Chroma Meter, Minolta CR‐200). At intervals after harvest (0, 6, 12, 24, 48, 96, and 120 h) the florets from each branchlet were removed with a sterile razor blade, frozen in liquid nitrogen and stored at –80 °C. In a second experiment, branchlets were excised from harvested broccoli heads and placed on moistened sterile pads inside plastic pillows. Gas mixes were pumped through the pillows (flow rate of 200 ml min–1) to give atmospheres containing 20% oxygen and <0.01% carbon dioxide (air), or 10% carbon dioxide and 5% oxygen (CA). Gas samples were taken from the pillows at daily intervals and analysed for oxygen and carbon dioxide content (Shimadzu GC‐9A, TCD detector). At intervals after harvest (0, 12, 24, 48, 72, 96, and 120 h) the florets from each branchlet were removed with a sterile razor blade, frozen in liquid nitrogen and stored at –80 °C. All experiments were carried out at 20 °C and the branchlets were held in the dark, except for when samples were taken. All treatments were replicated on three branchlets, and each replicate comprised the pooled floret tissue of one branchlet. Chlorophyll analysis Chlorophyll was extracted from fresh frozen broccoli tissue (225 mg) in 1 ml of N,N‐dimethyl formamide as previously described (Moran and Porath, 1980). The quantity of chlorophyll (mg g–1 FW) was calculated using extinction coefficients for chlorophyll a and b previously described by Inskeep and Bloom (1985). Endo‐protease activity Endo‐protease activity was determined using azo‐casein (Megazyme International Ireland Ltd, Co. Wicklow, Ireland) following the product protocols. Floret tissue (100 mg) was homogenized with 1 ml phosphate buffer (100 mM sodium phosphate, 30 mM cysteine, 30 mM EDTA, pH 7.0) and left for 2 h on ice with periodic vortexing for the extraction of proteases. Following centrifugation (14 000 g, 30 min, 4 °C), 0.5 ml supernatant was incubated for 24 h at 40 °C with an equal volume of 2% azo‐casein. The reaction was stopped by adding 3 ml trichloroacetic acid (5% w/v) and the absorbance of the supernatant was determined at 440 nm. Protease activity was calculated by reference to a papain regression equation. One protease unit is defined as the amount of enzyme required to hydrolyse (and TCA solubilize) 1 µmol of tyrosine equivalents min–1 from soluble casein under standard assay conditions (pH 7.0, 40 °C). Proteases extracted from broccoli florets 72 h after harvest were incubated with protease inhibitors, in order to determine the proportion of protease activity in senescing florets that could be attributed specifically to cysteine protease activity. Aliquots (50 µl) of inhibitors (leupeptin 25 mg ml–1 (Sigma Chemical Company, Milwaukee, IL, USA), iodoacetamide 118 mg ml–1 (Sigma), aprotinin 10 mg ml–1 (Roche Applied Science, Roche Diagnostics NZ Ltd, Auckland, New Zealand), and EDTA 10 mg ml–1 (BDH Chemicals Ltd, Poole, England) were added to 500 µl of extract (pH 7.0) and incubated for 21 h at 40 °C with an equal volume of 2% azo‐casein. The reaction was stopped by adding trichloroacetic acid and the absorbance of the supernatant was determined at 440 nm. Preparation of protein extracts Soluble proteins were extracted from broccoli tissue (50 mg) by grinding the tissue in 500 µl extraction buffer (100 mM Tris‐HCl, pH 7.6, 10 mM MgSO4, 10 mM dithiothreitol (Sigma)) and incubating on ice for 1 h with vortexing every 15 min. The samples were then centrifuged (10 min, 4 °C, 10 000 g) and the protein‐containing supernatant was retained. The protein concentration in the supernatant was determined using a Coomassie dye‐binding assay (Bio‐Rad Laboratories Ltd, Richmond, CA, USA) with bovine serum albumin (Sigma) as the standard. Substrate‐containing gel electrophoresis (zymography) Non‐denatured protein extracts containing equivalent amounts of protein (10 µg) were analysed by SDS‐PAGE on 10% acrylamide gels containing gelatin (Zymogram Ready gels, Bio‐Rad Laboratories Ltd) following procedures described previously (Eason et al., 2002). Isolation and characterization of broccoli cysteine protease cDNAs Cysteine protease cDNAs were isolated from a 48 h senescing broccoli floret cDNA library (Pogson et al., 1995) by screening the library with heterologous probes at 55°C. The selected heterologous probes were from Dianthuscaryophyllus (DCCP1, Jones et al., 1995), Hemerocallis (HSPSEN102, Guerrero et al., 1998), Brassicanapus (LSC790, Buchanan‐Wollaston and Ainsworth, 1997), and Sandersoniaaurantiaca (PRT5, Eason et al., 2002). Hybridizing plaques were taken through two rounds of purification. Plasmid DNA was isolated from Escherichiacoli using Wizard miniprep kits (Promega, Madison, WI, USA). DNA restriction analysis was performed using standard techniques (Sambrook et al., 1989). DNA inserts were purified from agarose gel slices using the Hi‐Pure PCR clean‐up kit (Roche, Mannheim, Germany). DNA sequencing was performed using the Applied Biosystems (Foster City, California, USA) Dye Primer Cycle Sequencing kits using the chain termination method (Sanger et al., 1977) in conjunction with an Applied Biosystems 373A DNA Sequencer at the University of Waikato DNA Sequencing Facility. The sequence programs of DNASTAR (Madison, WI, USA) were used to compile and align sequences. The programs BLASTN and BLASTP (Altschul et al., 1990) at the NCBI (Bethesda, Maryland, USA) web site were used to search the computer databases. The ExPASy (Geneva, Switzerland) web site (and programs contained within) was used to identify potential cleavage sites of signal peptides (Nielsen et al., 1997) and to calculate molecular masses and pIs of the deduced protein sequences. Southern and northern blot analysis Genomic DNA was isolated from broccoli floret tissue using the Plant DNA Isolation Kit (Roche, Mannheim, Germany) according to the manufacturer’s instructions. For Southern blot analysis, 10 µg of DNA from broccoli was digested with restriction enzymes, separated on 0.8% agarose gel and transferred overnight in 0.4 N NaOH on to Hybond N+ membrane (Amersham Biosciences). The Hind III digested λ DNA ladder (Life Technologies, Rockville, Maryland, USA) was used as a molecular weight marker. Probes were radio‐labelled using a random priming kit (Amersham Biosciences). The Southern blots were hybridized at 50 °C in hybridizing solution (Church and Gilbert, 1984). Low stringency washes were performed by gentle shaking in 2× SSC and 1× SSC for 15 min at 55 °C each (SSC solutions contained 0.1% (w/v) SDS). Radioactivity on the membrane was detected by autoradiography (Kodak Biomax MR film, –80 °C). After a 4 d exposure, the blots were washed at high stringency (0.1× SSC, 15 min, 65 °C) and re‐exposed for a further week. RNA was isolated from frozen ground floret tissue using Trizol Reagent (Life Technologies) according to the manufacturer’s instructions. For northern blot analysis, 10 µg of total RNA was electrophoresed on a 1.2% (w/v) agarose gel (1× MOPS, 8% (v/v) formaldehyde) and transferred overnight in 0.05 N NaOH to Hybond N+ membrane (Amersham Biosciences, Piscataway, New Jersey, USA). Prehybridization and hybridization was carried out at 65 °C using the following protocol for Hybond N+ membranes (Amersham Biosciences). Probes were radiolabelled using a random priming kit (Amersham Biosciences). Following hybridization, membranes were washed in 3× SSC, 0.1% (w/v) SDS (15 min); 1× SSC, 0.1% (w/v) SDS (15 min); and 0.1× SSC, 0.1% (w/v) SDS (2 washes of 15 min) at 65 °C, then exposed to Kodak Biomax MR film at –80 °C. RNA loadings were verified by ethidium bromide staining and rehybridization with a cDNA encoding the asparagus 25/26S rRNA (pTIP6; KM Davies, unpublished data). The northern blots were replicated, with each probe hybridized to two replicate sets of blots (e.g. air/water/sucrose or air/CA) with similar results. Statistical analysis Data were statistically analysed by analysis of variance (Minitab 13) and least significant differences (LSD, 5% level) were calculated. Results Water loss from broccoli tissue after harvest Broccoli tissue senesces rapidly after harvest and one of the first signs of senescence is a loss of turgor and wilting. Branchlets held in air at 20 °C lost over 4% of their initial weight within 4 h. They lost a total of 45% weight by 96 h (Fig. 1A). The majority of this water loss occurs from the florets. By coating different surfaces with petroleum jelly it was determined that around 70% of the water lost over a 4 d period was lost from the florets; 15% was lost from the stem surface and 15% through the cut end (Fig. 1A). When branchlets were held in air, the visual quality was retained when the florets were coated with petroleum jelly. Anatomical examination confirmed that broccoli main stems have scattered stomata, approximately 5–25 mm–2. Sepals by contrast have a high stomatal density (250–300 mm–2) and, of course, a huge surface area. Together these account for the major route of water loss being the florets. Prior treatment of branchlets with ABA to close the stomata was less effective than coating the florets with petroleum jelly, both in terms of reducing water loss (Table 1) and retaining visual quality. It is evident that broccoli stems act as a capacitor: rapid water loss from florets can occur faster than water can be drawn from the stems, making the florets and subtending stemlets very prone to short‐term water stress, but this will be slowly rectified as water moves into the florets from the stems, if free water is supplied at the cut end or transpiration is reduced. Holding broccoli branchlets in water leads to an increase in fresh weight after harvest; this was variable between experiments, but was in the range of 200 mg h–1 for the first 6 h (Fig. 1B), after which time the weight remained fairly stable. When 70 mM sucrose was added to the water as an energy source, branchlet growth was reduced to 55 mg h–1 for the first 6 h (Fig. 1B), and net water loss began soon afterwards. Water loss reached only 6% of the initial weight after 70 h for sucrose fed branchlets (Fig. 1B), compared with around 37% for branchlets held in air (Fig. 1A). MRI images showed non‐uniform water loss from the tissues. There was a strong reduction in the MRI signal from broccoli florets 26 h after harvest (Fig. 2A) compared with images 2 h after harvest (Fig. 2B), indicating that the florets in particular are experiencing dehydration as a result of water loss. Protease activity during senescence The progress of broccoli floret senescence can be followed by measuring colour change as tissue undergoes chlorosis. The colour of the floret sepals changes from green to yellow, which correlates with a measurable decrease in hue angle (Fig. 3A). Within 48 h, branchlets that were held in air at 20 °C were showing signs of chlorosis (decline in hue angle, Fig. 3A). The floret tissue continues to yellow and dehydrate, becoming a darker purple/brown at 96 h, which is measured here as an increase in the hue angle (Fig. 3A). Broccoli branchlets that were held in water showed the same level of chlorosis as those held dry, but were firmer to the touch throughout the post‐harvest period. Branchlets that were held in sucrose solution (2% w/v) retained their green colour for longer and chlorosis, measured as a decrease in hue angle, was less severe (Fig. 3A). Endo‐protease activity in broccoli floret extracts was determined using azo‐casein. Protease activity increased linearly after 48 h in tissue that had been held dry or in water (Fig. 3B). Sucrose treatment delayed the elevation of protease activity in floret tissue by 24 h, with endo‐protease activity of sucrose‐treated tissue increasing 72 h after harvest (Fig. 3B). Zymography combines separation by electrophoresis with direct enzyme activity. Following insitu renaturation of the electrophoresed proteins, incubation in activity buffer (pH 7.5) and staining of the remaining gelatin substrate, protein bands with proteolytic activity were visualized as clear bands on a dark background (Fig. 4A). Multiple protein bands exhibiting protease activity were detected in the extracts from broccoli florets, ranging in size from c. 45–70 kDa (Fig. 4A). The activity of one of the protein bands was present throughout the post‐harvest period, increasing during the later stages of senescence, while the protease activity of the other three bands increased from 72 h. The increase in protease activity mirrors a decline in soluble protein content of broccoli tissue after harvest. Zymographic analysis revealed an increase in activity of one protease‐active band from harvest, and azocasein assays show a rise in protease activity from 24 h after harvest (Figs 3B, 4A). Protein content of floret tissue decreases from 10 mg g–1 (FW) at harvest, to 6 mg g–1 (FW) at 72 h (Fig. 4B), then less rapidly to 5.2 mg g–1 at 120 h. The senescence of broccoli heads held in low oxygen, high carbon dioxide atmospheres is substantially reduced, with no significant decline in the total chlorophyll content of broccoli florets after 96 h compared to a rapid decline from 24 h in the chlorophyll content of tissue held in air (Fig. 5A). The protein content of CA stored broccoli is not significantly different for the first 120 h after harvest, compared to a decline in soluble protein in air‐stored broccoli from 6 h (Fig. 5B). The endo‐protease activity of CA‐stored tissue does not increase above a baseline level for the duration of the trial, whereas the protease activity in control tissue (held in air) increased dramatically between 24 h and 120 h (Fig. 5C). The specificity of the proteases isolated from floret tissue at 72 h after harvest (in air, at 20 °C) was determined by incubating extracts with aliquots of protease inhibitors. A range of chemicals with known protease inhibitor action was used to determine the class of proteases that had a role in broccoli floret senescence. The chemicals used are known specifically to inhibit cysteine proteases (iodoacetamide), serine proteases (aprotinin), both cysteine and serine proteases (leupeptin), and metalloproteases (EDTA). Leupeptin and iodoacetamide, known to inhibit cysteine proteases, reduced the endo‐protease activity by 44% and 14%, respectively (Fig. 6). In addition, aprotinin, which inhibits serine proteases, reduced endo‐protease activity by 10% (Fig. 6). Isolation and characterization of broccoli cysteine protease cDNAs Four cysteine protease cDNAs (BoCP1, BoCP2, BoCP3, BoCP4, Table 2) were isolated from a 48 h broccoli floret cDNA library after screening with heterologous cysteine protease cDNAs. Sequence analysis confirmed that all of the broccoli cDNAs isolated here encoded putative cysteine proteases. However, the isolated cDNAs were not homologues with the genes that were used to isolate them, nor were they highly similar to previously identified broccoli cysteine proteases. The four broccoli cDNAs, have high similarity to RD21 and RD19, two genes that encode drought‐inducible cysteine proteases in Arabidopsisthaliana (Table 2). Phylogenetic analysis indicated that BoCP2 and BoCP3 (which share 76.3% identity) are most closely related to RD21 and NP568620 (Arabidopsisthaliana, uncharacterized). BoCP1 shares 55%, 53% and 29% identity to BoCP2, BoCP3 and BoCP4, respectively, and is most similar to another uncharacterized Arabidopsis gene (NP566633, unpublished data). BoCP4, at less than 35% identity to the other BoCP cDNAs isolated here, is most similar to the Arabidopsis RD19. Genomic Southern analysis revealed that each putative cysteine protease gene is a member of a small gene family (data not shown). BoCP2 and BoCP3 have the most similar nucleotide sequences of the four isolated cDNAs, with 76.3% identity. Under stringent hybridization conditions, however, no common bands were found between the Southern blots for these two cDNAs, indicating that the cDNAs are not cross hybridizing. Transcripts of the four putative cysteine protease genes accumulate during floret senescence (Fig. 7). The mRNA for BoCP1 is present at harvest and increases in abundance in the floret tissue of broccoli branchlets that are held in air, to a maximum at 48 h. BoCP2 is also expressed at low levels at harvest and accumulates from 24 h to a maximum level at 48–96 h after harvest. BoCP3 is detected from 6 h after harvest, and mRNA levels accumulate from 24 h to a maximum at 48–72 h after harvest. The pattern of expression for BoCP4 differs from the other three cDNAs with mRNA accumulating from 12 h after harvest until 24 h, after which time the levels decline (Fig. 7). In order to learn more about the expression of these putative cysteine protease genes during floret senescence, the expression of the genes was analysed following treatments that delay broccoli floret senescence (water, sucrose). Wilting is prevented when harvested branchlets are held in water, and chlorosis is further delayed when harvested branchlets are held in solutions of sucrose (2% w/v, Fig. 3A). The expression of all four putative cysteine proteases was delayed when branchlets were held in water (Fig. 7). For water‐treated broccoli, mRNA levels for BoCP1 began to accumulate from 6 h after harvest, rather than immediately after harvest, as was the case for tissue held in air. BoCP2 mRNA accumulated from 24 h after harvest in floret tissue of broccoli held in water, and did not accumulate to the same levels seen in floret tissue that had been held in air. BoCP3 mRNA was less abundant at 24 h for tissue held in water and reached maximum expression levels at 96 h after harvest, 48 h later than the maximum expression of air‐treated tissue. BoCP4 mRNA was expressed at very low levels in tissue that had been held in water reaching a maximum at 96 h compared to a maximum at 24 h for tissue held in air. Broccoli branchlets that were held in sucrose had further reduced expression of BoCP1 and BoCP3, but both BoCP2 and BoCP4 were expressed at similar levels to those of tissue that had been held in water. Controlled atmosphere storage of broccoli has a major impact on the senescence of the tissue. Broccoli held under high carbon dioxide/low oxygen atmospheres (10% CO2, 5% O2) does not become chlorotic (Fig. 5A), lose soluble protein (Fig. 5B) or have a senescence‐associated increase in protease activity (Fig. 5C) for the first 120 h after harvest. In addition, expression of cysteine protease genes is reduced. Transcripts for BoCP1 and BoCP3 did not accumulate in CA‐treated broccoli for the first 96 h after harvest (Fig. 8). Both BoCP2 and BoCP4 were expressed in the CA‐treated broccoli, but the level of mRNA accumulation for these two genes was lower than for the air controls (Fig. 8). Discussion The senescence of broccoli is induced artificially through the act of harvest, which causes the stressed tissue to undergo a series of events that closely resembles the natural senescence of leaves and flowers (e.g. drop in cellular respiration, ethylene production, loss of chlorophyll, carbohydrates and proteins). One of the earliest visible signs of senescence in harvested broccoli tissues is wilting, a consequence of the water stress encountered in the harvested tissue. Branchlets held in air at 20 °C were measurably lighter through water loss within 4 h of harvest. In the present investigation, 70% of the water lost from broccoli branchlets was lost via the floret tissue. Coating floret tissue with petroleum jelly was more effective than ABA treatments in reducing water loss (Table 1), indicating that water was lost both through stomata and across the epidermis. Visual quality of tissues (e.g. healthy green sepals) was retained for longer in tissues with reduced water loss. The ability to delay the onset of senescence by reducing water loss indicates that one of the signals initiating death in the cells of harvested broccoli florets may be a response to water stress. The regulation of senescence in the current situation, however, is more complex. These data show that senescence of branchlets treated with sucrose was delayed (chlorosis was reduced Fig. 3A). Water uptake was also reduced in sucrose‐treated branchlets (55 mg h–1 for the first 6 h, instead of 200 mg h–1 for branchlets held in water). After 96 h, sucrose‐treated branchlets lost water at a rate significantly greater than branchlets held in water. Therefore, although broccoli that was treated with sucrose took longer to become chlorotic, it would have been under more water stress by 96 h after harvest than broccoli held in water. This suggests that although immediate water stress may be a harvest signal, the later yellowing is not a simple consequence of this water stress. Further investigation of how pre‐senescent cells perceive these water loss/stress signals may provide some insight into the initiation of harvest‐induced senescence. Indeed, the act of harvest exposes the tissues of broccoli heads to many serious stresses (including wounding and subsequent ethylene production, turgor/osmotic changes, halted nutrient supply, changing hormone status, sucrose loss), any one of which may initiate senescence and lead to cell death. The series of events that links the signals initiating senescence to the series of programmed events that cause the death of cells are now being unravelled by plant researchers. Cysteine proteases are postulated to have a role in the process (Buchanan‐Wollaston and Ainsworth, 1997; Noodén et al., 1997; Schmid et al., 1999) with several senescing plant tissues expressing papain‐type cysteine endopeptidases (Beers et al., 2000). The protease activity of broccoli floret tissue that undergoes premature harvest‐induced senescence is characterized here. This study shows that floret endo‐protease activity increased from 24 h after harvest (Figs 3B, 5C), during which time chlorophyll degradation was initiated (Fig. 5A), sepal tissue became chlorotic (Fig. 3A), and protein levels declined (Figs 4B, 5B). Protease‐active protein bands, visualized as clear bands on a zymogram, were present in the floret tissue at harvest and increased in abundance as senescence progressed, particularly at 72 h when new protease bands were observed (Fig. 4A). These data are consistent with the elevation in endo‐protease activity measured by azo‐casein assays (Figs 3B, 5C). Both sucrose treatment and controlled atmospheric storage are known to delay the onset of senescence in harvested tissues (Irving and Joyce, 1995; Hurst et al., 1997). These treatments also delayed the elevation in protease activity associated with floret senescence in broccoli (Figs 3B, 5C), further emphasizing the important role proteolysis plays during harvest‐induced senescence. The use of protease inhibitors invitro showed that the protease activity at 72 h after harvest was due in part to specific cysteine proteases. The differential inhibition of activity by the chemical inhibitors is a factor of their concentration and specificity. For instance, leupeptin inhibits both cysteine and serine proteases and had the largest impact on invitro protease activity, reducing endo‐protease activity by 44%. Chemical inhibition with iodoacetamide and aprotinin, which specifically inhibit the smaller pools of cysteine or serine proteases, indicate that at least 14% of the invitro protease activity in broccoli tissues is due to papain‐like cysteine proteases and 10% is due to serine proteases (Fig. 6). In this paper, we show that at least eight distinct cDNAs encoding putative cysteine proteases are induced differentially during senescence in broccoli florets (four isolated here: BoCP1, BoCP2, BoCP3, BoCP4; and four that have been shown by other researchers to cross hybridize with B. oleracea mRNA: LSC7, LSC790, LSC833, SAG12; Page et al., 2001). The four cDNAs characterized in the current research have high similarity to dehydration‐responsive cysteine proteases isolated from Arabidopsis (RD19, RD21). The mRNAs for BoCP1, BoCP4, BoCP3, and BoCP2 accumulate from 6, 12, 24, and 24 h after harvest respectively (Fig. 7) with maximum expression at 24–48, 24–48, 48–72, and 48–96 h after harvest, respectively, for tissue held in air at 20 °C. The induction of these putative cysteine protease genes, in particular BoCP1, is a relatively early event, concomitant with water loss and prior to chlorosis. As would be expected, the induction of these genes also precedes the increase in protease activity measured in senescing floret tissue (compare Figs 3B/7 and 5C/8). Treatments that delay the onset of senescence in broccoli (e.g. water and sucrose treatment and CA storage) also delay the induction of the cysteine protease mRNAs, but to different extents. For instance, branchlets that were held in water had lower levels of mRNA accumulation for all four cDNAs than the air controls, but BoCP2 and BoCP4 were the most dramatically affected, with mRNA for BoCP4 accumulating to maximum levels by 96 h compared to 12 h in air (Fig. 7). Sucrose treatment of branchlets further delayed the accumulation of BoCP1 and BoCP3 in florets compared to the water controls, but did not further influence the expression of BoCP2 and BoCP4 over that of the water‐treated tissues (Fig. 7). Controlled atmosphere storage had the largest impact on cysteine protease expression. Messenger RNA for BoCP1 and BoCP3 did not accumulate in the floret tissue of branchlets that were stored in CA during the first 96 h after harvest, and the accumulation of BoCP2 and BoCP4 mRNAs was reduced (Fig. 8). The differential expression patterns of these four cDNAs indicate that at least three separate signals may regulate the expression of cysteine proteases during harvest‐induced senescence in broccoli. Firstly, the cDNAs are induced earlier after harvest in wilted as opposed to turgid tissue, suggesting that a turgor or osmotic signal may regulate gene expression in a manner similar to that suggested for Arabidopsis RD19 and RD21 cDNAs (Koizumi et al., 1993). Secondly, supplying additional sucrose further delayed the induction of BoCP1 and BoCP3. Perhaps sugar signalling is effective here, indeed it is well documented in the regulation of gene expression in other senescing systems (Sheen et al., 1999). Thirdly, high CO2, low O2 prevented the accumulation of mRNAs for BoCP1 and BoCP3 for at least 96 h. The cellular mechanisms that effect delayed senescence through CA storage are not understood (Kader and Ben‐Yehoshua, 2000), but, as is shown here, are extremely effective. Further investigations are being carried out into the regulation of senescence imposed by this type of post‐harvest storage, in particular to determine the component of the gas mix that is most effective, high CO2 or low O2 (R Lill, B Page, unpublished results). RD21 and RD19 were isolated from Arabidopsis during a search for genes that were responsive to desiccation (Yamaguchi‐Shinozaki et al., 1992). The researchers found that the mRNA for RD21 and RD19 accumulated following drought and salt stress, but not after ABA, gibberellic acid, cold or heat treatments. Further study showed RD21 contained a C‐terminal extension sequence composed of a 2 kDa proline‐rich domain and a 10 kDa domain homologous to the domains of granulin family proteins (Yamada et al., 2001). The RD21 protein is produced as a pre‐pro‐protein (58 kDa), which is processed via an intermediate (38 kDa) into a mature protein (33 kDa) through the removal of the pro‐peptide followed by removal of the granulin domain (Yamada et al., 2001). The RD21 precursor (38 kDa) accumulates specifically in endoplasmic reticulum (ER) bodies of the epidermal cells in Arabidopsis leaves, and after stress (e.g. salt) the ER bodies fuse with the vacuole, delivering the precursor directly into the vacuole where it becomes active (33 kDa, Hayashi et al., 2001). A search of the GenBank database revealed that the broccoli cDNAs, BoCP2 and BoCP3, are most similar to RD21. However, cDNAs of RD21 homologues have also been isolated from Oryzasativa (D90406, Watanabe et al., 1991), Pisumsativum (X66061, Granell et al., 1992) and Solanumtuberosum (AJ245924, Avrova et al., 1999), where the mRNAs are induced by gibberellin, pathogen attack and during senescence. In the current study, the expression of the broccoli RD21 homologues is elevated in tissues under water stress and affected by supply of sucrose and CA storage. The expression of dehydration‐responsive cysteine proteases in broccoli florets after harvest suggests that one of the stresses that triggers cell death following harvest for this plant is water stress. It has been shown that floret tissue is particularly vulnerable to water loss (Fig. 2), and net water loss began within 4 h for branchlets held in air, and reached 35% of initial fresh weight after 72 h. The induction of mRNAs for four putative cysteine proteases is delayed as a direct result of providing the tissue with an external source of water. Although the use of controlled atmosphere storage is known to prevent senescence in a range of harvested produce (Kader and Ben‐Yehoshua, 2000), the underlying scientific basis for why such a technology should work has not been determined. It is shown here that one of the biochemical changes enforced through CA storage of broccoli is the prevention and reduction of expression of specific cysteine protease genes, and the subsequent reduction of proteolytic action. Acknowledgements We thank Michelle Jones (Ohio State University) for her gift of DCCP1, Mike Reid (University of California) for his gift of HSPSEN102, and Vicky Buchanan‐Wollaston (HRI Wellesbourne, UK) for her gift of LSC790. We thank Chris Clark (HortResearch) for his MRI analysis and Tatyana Pinkney and Lindsay Greer for technical assistance in water loss experiments. This work was supported by a grant from the Foundation for Research, Science and Technology, New Zealand. View largeDownload slide Fig. 1. Weight change of broccoli branchlets. (A) Broccoli branchlets were held in water (filled squares) or air at 20 °C. Air‐stored branchlets were coated with paraffin jelly to cover florets (open squares), stem and cut end (filled triangles), cut end alone (open triangles, or left uncoated (filled circles). (B) Broccoli branchlets in water (filled squares) or 70 mM sucrose (open squares) at 20 °C. Data are significantly different at each time point after harvest. View largeDownload slide Fig. 1. Weight change of broccoli branchlets. (A) Broccoli branchlets were held in water (filled squares) or air at 20 °C. Air‐stored branchlets were coated with paraffin jelly to cover florets (open squares), stem and cut end (filled triangles), cut end alone (open triangles, or left uncoated (filled circles). (B) Broccoli branchlets in water (filled squares) or 70 mM sucrose (open squares) at 20 °C. Data are significantly different at each time point after harvest. View largeDownload slide Fig. 2. Magnetic resonance image (MRI) of broccoli after harvest. MRI pseudo‐proton density images using Tr=8000 ms, Te=25 ms of (A) longitudinal section through a broccoli branchlet at 2 h, and (B) longitudinal section through a broccoli branchlet at 26 h after harvest. Bright signals indicate high water activity. View largeDownload slide Fig. 2. Magnetic resonance image (MRI) of broccoli after harvest. MRI pseudo‐proton density images using Tr=8000 ms, Te=25 ms of (A) longitudinal section through a broccoli branchlet at 2 h, and (B) longitudinal section through a broccoli branchlet at 26 h after harvest. Bright signals indicate high water activity. View largeDownload slide Fig. 3. Colour change (A) and endo‐protease enzyme activity (B) in broccoli floret tissue after harvest. Broccoli was held in air (filled triangles), solutions of water (filled squares) or sucrose (open squares; 2% w/v) for 0–96 h at 20 °C after harvest. View largeDownload slide Fig. 3. Colour change (A) and endo‐protease enzyme activity (B) in broccoli floret tissue after harvest. Broccoli was held in air (filled triangles), solutions of water (filled squares) or sucrose (open squares; 2% w/v) for 0–96 h at 20 °C after harvest. View largeDownload slide Fig. 4. Protease activity in broccoli floret tissue after harvest. (A) Substrate‐containing gel electrophoresis of proteins extracted from broccoli floret tissues held at 20 °C for 0–120 h after harvest. Proteins demonstrating proteolytic activity are seen as clear bands against a dark background (indicated by arrows). (B) Protein content of broccoli tissue after harvest. View largeDownload slide Fig. 4. Protease activity in broccoli floret tissue after harvest. (A) Substrate‐containing gel electrophoresis of proteins extracted from broccoli floret tissues held at 20 °C for 0–120 h after harvest. Proteins demonstrating proteolytic activity are seen as clear bands against a dark background (indicated by arrows). (B) Protein content of broccoli tissue after harvest. View largeDownload slide Fig. 5. Chlorophyll content (A), soluble protein content (B) and endo‐protease enzyme activity (C) in broccoli floret tissue stored in controlled atmospheres. Broccoli was held in humidified air (filled squares) or controlled atmosphere (open squares; 10% CO2, 5% O2) for 0–120 h at 20 °C. View largeDownload slide Fig. 5. Chlorophyll content (A), soluble protein content (B) and endo‐protease enzyme activity (C) in broccoli floret tissue stored in controlled atmospheres. Broccoli was held in humidified air (filled squares) or controlled atmosphere (open squares; 10% CO2, 5% O2) for 0–120 h at 20 °C. View largeDownload slide Fig. 6. Inhibition of endo‐protease activity in broccoli floret tissue. Protein extracts were incubated with water‐soluble protease inhibitors (black bars). Protease activity was determined by hydrolysis of azo‐casein at pH 7, 40 °C. Values are the means of three replicates, each replicate was extracted from floret tissue shaved from one branchlet. LSD (5%)=11.8 (24 df). View largeDownload slide Fig. 6. Inhibition of endo‐protease activity in broccoli floret tissue. Protein extracts were incubated with water‐soluble protease inhibitors (black bars). Protease activity was determined by hydrolysis of azo‐casein at pH 7, 40 °C. Values are the means of three replicates, each replicate was extracted from floret tissue shaved from one branchlet. LSD (5%)=11.8 (24 df). View largeDownload slide Fig. 7. Expression of putative broccoli cysteine proteases genes during senescence. RNA was extracted from the floret tissues of broccoli branchlets that had been held dry, in water or in sucrose (2% w/v) for 0, 6, 12, 24, 48, 72, and 96 h after harvest. Isolated RNA was separated by electrophoresis in formaldehyde‐containing 1.3% agarose gels. RNA loadings were visualized by hybridization with pTIP6 (asparagus 25/26S rRNA). View largeDownload slide Fig. 7. Expression of putative broccoli cysteine proteases genes during senescence. RNA was extracted from the floret tissues of broccoli branchlets that had been held dry, in water or in sucrose (2% w/v) for 0, 6, 12, 24, 48, 72, and 96 h after harvest. Isolated RNA was separated by electrophoresis in formaldehyde‐containing 1.3% agarose gels. RNA loadings were visualized by hybridization with pTIP6 (asparagus 25/26S rRNA). View largeDownload slide Fig. 8. Expression of putative broccoli cysteine proteases genes during controlled atmosphere storage. RNA was extracted from the floret tissues of broccoli branchlets that had been held in air or controlled atmosphere (10% CO2, 5% O2) for 0, 12, 24, 48, 72, and 96 h. Isolated RNA was separated by electrophoresis in formaldehyde‐containing 1.3% agarose gels. RNA loadings were visualized by hybridization with pTIP6 (asparagus 25/26S rRNA). View largeDownload slide Fig. 8. Expression of putative broccoli cysteine proteases genes during controlled atmosphere storage. RNA was extracted from the floret tissues of broccoli branchlets that had been held in air or controlled atmosphere (10% CO2, 5% O2) for 0, 12, 24, 48, 72, and 96 h. Isolated RNA was separated by electrophoresis in formaldehyde‐containing 1.3% agarose gels. RNA loadings were visualized by hybridization with pTIP6 (asparagus 25/26S rRNA). Table 1. Effect of 22 h pretreatment on the rate of water loss from broccoli branchlets in air at 20 °C   Treatment–Air  LSD (5%)  Pretreatment  Water  ABA (0.2 mM)  ABA (2 mM)  Petroleum jelly    Water loss (first 8 h)  (mg h–1)  168  144  133  98  18    Treatment–Air  LSD (5%)  Pretreatment  Water  ABA (0.2 mM)  ABA (2 mM)  Petroleum jelly    Water loss (first 8 h)  (mg h–1)  168  144  133  98  18  View Large Table 2. Broccoli dehydration‐responsive cysteine protease cDNAs CDNA name  Accession number  cDNA length (bp)  Predicted protein (aa)  Mr of predicted protein (kDa)  Identitya  Reference  BoCP1b  AF454956  1582  445  44.75  83% NP566633 Arabidopsis  Unpublished            75% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP2  AF454957  1699  460  50.61  83% NP568620 Arabidopsis  Unpublished            73% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP3  AF454958  1830  485  53.79  80% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP4  AF454959  1253  686  40.21  89% P43296 (RD19) Arabidopsis  Koizumi et al., 1993  CDNA name  Accession number  cDNA length (bp)  Predicted protein (aa)  Mr of predicted protein (kDa)  Identitya  Reference  BoCP1b  AF454956  1582  445  44.75  83% NP566633 Arabidopsis  Unpublished            75% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP2  AF454957  1699  460  50.61  83% NP568620 Arabidopsis  Unpublished            73% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP3  AF454958  1830  485  53.79  80% P43297 (RD21) Arabidopsis  Koizumi et al., 1993  BoCP4  AF454959  1253  686  40.21  89% P43296 (RD19) Arabidopsis  Koizumi et al., 1993  a Percentage similarity at amino acid level. b BoCP1 is a partial cDNA (short c. 5 amino acids). View Large References AltschulS, Gish W, Miller W, Myers E, Lipman D. 1990. Basic local alignment search tool. Journal of Molecular Biology  215, 403–410. Google Scholar AvrovaAO, Stewart HE, De Jong WD, Heilbronn J, Lyon GD, Birch PR. 1999. A cysteine protease gene is expressed early in resistant potato interactions with Phytophthora infestans. Molecular Plant–Microbe Interactions  12, 1114–1119. Google Scholar BeersEP, Woffenden BJ, Zhao CS. 2000. Plant proteolytic enzymes: possible roles during programmed cell death. Plant Molecular Biology  44, 399–415. Google Scholar Buchanan‐WollastonV, Ainsworth C. 1997. Leaf senescence in Brassica napus—cloning of senescence related genes by subtractive hybridization. Plant Molecular Biology  33, 821–834. Google Scholar CallisJ, Vierstra RD. 2000. Protein degradation in signalling. Current Opinion in Plant Biology  3, 381–386. Google Scholar CervantesE, Rodriguez A, Nicolas G. 1994. Ethylene regulates the expression of a cysteine proteinase gene during germination of chickpea (Cicer arietinum L.). Plant Molecular Biology  25, 207–215. Google Scholar ChevalierC, Bourgeois E, Pradet A, Raymond P. 1995. Molecular cloning and characterization of six cDNAs expressed during glucose starvation in excised maize (Zea mays L) root tips. Plant Molecular Biology  28, 473–485. Google Scholar ChurchG, Gilbert W. 1984. Genomic sequencing. Proceedings of the National Academy of Sciences, USA  81, 1991–1995. Google Scholar ClarkCJ, Drummond LN, MacFall JS. 1998. Quantitative magnetic resonance imaging of kiwifruit (Actinidia deliciosa) during growth and ripening. Journal of the Science of Food and Agriculture   78, 349–358. Google Scholar ClarkeSF, Jameson PE, Downs C. 1994. The influence of 6‐benzylaminopurine on post‐harvest senescence of floral tissues of broccoli (Brassica oleracea var. Italica). Plant Growth Regulation   14, 21–27. Google Scholar DownsCG, Somerfield SD. 1997. Asparagine synthetase gene expression increases as sucrose declines in broccoli after harvest. New Zealand Journal of Crop and Horticultural Science   25, 191–195. Google Scholar EasonJR, Ryan DJ, Pinkney TT, O’Donoghue EM. 2002. Programmed cell death during flower senescence: Isolation and characterization of cysteine proteinases from Sandersonia aurantiaca. Functional Plant Biology   29, 1055–1064. Google Scholar GranellA, Harris N, Pisabarro AG, Carbonell J. 1992. Temporal and spatial expression of a thiolprotease gene during pea ovary senescence, and its regulation by gibberellin. The Plant Journal  2, 907–915. Google Scholar GuerreroC, Delacalle M, Reid MS, Valpuesta V. 1998. Analysis of the expression of two thiolprotease genes from daylily (Hemerocallis spp.) during flower senescence. Plant Molecular Biology  36, 565–571. Google Scholar HayashiY, Yamada K, Shimada T, Matsushima R, Nishizawa N, Hara‐Nishimura I. 2001. A proteinase‐storing body that prepares for cell death or stress in the epidermal cells of Arabidopsis. Plant Physiology  127, 1626–1634. Google Scholar HurstPL, Cheer V, Sinclair BK, Irving DE. 1997. Biochemcial responses of asparagus to controlled atmosphere storage at 20 °C. Journal of Food Biochemistry  20, 463–472 Google Scholar InskeepWP, Bloom PR. 1985. Extinction coefficients of chlorophyll a and b in N,N‐dimethylformamide and 80% acetone. Plant Physiology  77, 483–485. Google Scholar IrvingDE, Joyce DC. 1995. Sucrose supply can increase longevity of broccoli (Brassica oleracea) branchlets kept at 22 °C. Plant Growth Regulation   17, 251–256. Google Scholar JohnI, Hackett R, Cooper W, Drake R, Farrell A, Grierson D. 1997. Cloning and characterization of tomato leaf senescence‐related cDNAs. Plant Molecular Biology  33, 641–651. Google Scholar JonesJT, Mullet JE. 1995. A salt‐ and dehydration‐inducible pea gene, Cyp15a, encodes a cell‐wall protein with sequence similarity to cysteine proteases. Plant Molecular Biology  28, 1055–1065. Google Scholar JonesML, Larsen PB, Woodson WR. 1995. Ethylene‐regulated expression of a carnation cysteine proteinase during flower petal senescence. Plant Molecular Biology  28, 505–512. Google Scholar KaderA, Ben‐Yehoshua S. 2000. Effects of superatmospheric oxygen levels on post‐harvest physiology and quality of fresh fruits and vegetables. Post‐harvest Biology and Technology   20, 1–13. Google Scholar KardailskyIV, Brewin NJ. 1996. Expression of cysteine protease genes in pea nodule development and senescence. Molecular Plant–Microbe Interactions  9, 689–695. Google Scholar KingGA, Morris SC. 1994a. Early compositional changes during post‐harvest senescence of broccoli. Journal of the American Society for Horticultural Science  119, 1000–1005. Google Scholar KingGA, Morris SC. 1994b. Physiological changes of broccoli during early post‐harvest senescence and through the preharvest–postharvest continuum. Journal of the American Society for Horticultural Science  119, 270–275. Google Scholar KoizumiM, Yamaguchi‐Shinozaki K, Tsuji H, Shinozaki K. 1993. Structure and expression of two genes that encode distinct drought‐inducible cysteine proteinases in Arabidopsis thaliana. Gene  129, 175–182. Google Scholar LinthorstHJ, Van Der Does C, Brederode FT, Bol JF. 1993. Circadian expression and induction by wounding of tobacco genes for cysteine proteinase. Plant Molecular Biology  21, 685–694. Google Scholar LohmanKN, Gan S, John MC, Amasino RM. 1994. Molecular analysis of natural leaf senescence in Arabidopsis thaliana. Physiologia Plantarum  92, 322–328. Google Scholar MoranR, Porath D. 1980. Chlorophyll determination in intact tissues using N,N‐dimethylformamide. Plant Physiology  65, 478–479. Google Scholar NielsenH, Engelbrecht J, Brunak S, Vonheijne G. 1997. Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering   10, 1–6. Google Scholar NoodénLD, Guiamet JJ, John I. 1997. Senescence mechanisms. Physiologia Plantarum  101, 746–753. Google Scholar PageT, Griffiths G, Buchanan‐Wollaston V. 2001. Molecular and biochemical characterization of post‐harvest senescence in broccoli. Plant Physiology  125, 718–727. Google Scholar PogsonBJ, Downs CG, Davies KM. 1995. Differential expression of two 1‐aminocyclopropane‐1‐carboxylic acid oxidase genes in broccoli after harvest. Plant Physiology  108, 651–657. Google Scholar SambrookJ, Fritsch EF, Maniatis T. 1989. Molecular cloning: a laboratory manual. New York: Cold Spring Harbor Laboratory Press. Google Scholar SangerF, Nicklen S, Coulson A. 1977. DNA sequencing with chain terminating inhibitors. Proceedings of the National Academy of Sciences, USA  74, 5463–5467. Google Scholar SchafferMA, Fischer RL. 1988. Analysis of mRNAs that accumulate in response to low temperature identifies a thiol protease gene in tomato. Plant Physiology  87, 431–436. Google Scholar SchmidM, Simpson D, Gietl C. 1999. Programmed cell death in castor bean endosperm is associated with the accumulation and release of a cysteine endopeptidase from ricinosomes. Proceedings of the National Academy of Sciences, USA  96, 14159–14164. Google Scholar SheenJ, Zhou L, Jang JC. 1999. Sugars as signalling molecules. Current Opinion in Plant Biology  2, 410–418. Google Scholar SolomonM, Belenghi B, Delledonne M, Menachem E, Levine A. 1999. The involvement of cysteine proteases and protease inhibitor genes in the regulation of programmed cell death in plants. The Plant Cell  11, 431–443. Google Scholar StephensonP, Rubinstein B. 1998. Characterization of proteolytic activity during senescence in daylilies. Physiologia Plantarum  104, 463–473. Google Scholar VierstraRD. 1996. Proteolysis in plants: mechanisms and functions. Plant Molecular Biology  32, 275–302. Google Scholar WagstaffC, Leverentz MK, Griffiths G, Thomas B, Chanasut U, Stead AD, Rogers HJ. 2002. Cysteine protease gene expression and proteolytic activity during senescence of Alstroemeria petals. Journal of Experimental Botany  53, 233–240. Google Scholar WatanabeH, Abe K, Emori Y, Hosoyama H, Arai S. 1991. Molecular cloning and gibberellin‐induced expression of multiple cysteine proteinases of rice seeds (oryzains). Journal of Biological Chemistry  266, 16897–16902. Google Scholar WilliamsJ, Bulman M, Huttly A, Phillips A, Neill S. 1994. Characterization of a cDNA from Arabidopsis thaliana encoding a potential thiol protease whose expression is induced independently by wilting and abscisic acid. Plant Molecular Biology  25, 259–270. Google Scholar YamadaK, Matsushima R, Nishimura M, Hara‐Nishimura I. 2001. A slow maturation of a cysteine protease with a granulin domain in the vacuoles of senescing Arabidopsis leaves. Plant Physiology  127, 1626–1634. Google Scholar Yamaguchi‐ShinozakiK, Koizumi M, Urao S, Shinozaki K. 1992. Molecular cloning and characterization of 9 cDNAs for genes that are responsive to desiccation in Arabidopsis thaliana: sequence analysis of one cDNA clone that encodes a putative transmembrane channel protein. Plant and Cell Physiology  33, 217–224. Google Scholar
TI - Identification of dehydration‐responsive cysteine proteases during post‐harvest senescence of broccoli florets
JF - Journal of Experimental Botany
DO - 10.1093/jxb/erg105
DA - 2003-03-01
UR - https://www.deepdyve.com/lp/oxford-university-press/identification-of-dehydration-responsive-cysteine-proteases-during-L2C3QCf07S
SP - 1045
EP - 1056
VL - 54
IS - 384
DP - DeepDyve
ER -