WO1999024588A1

WO1999024588A1 - cDNA CLONE OF ENDO-β-MANNANASE FROM PLANT TISSUES

Info

Publication number: WO1999024588A1
Application number: PCT/CA1998/001034
Authority: WO
Inventors: J. Derek Bewley
Original assignee: University of Guelph
Current assignee: University of Guelph
Priority date: 1997-11-07
Filing date: 1998-11-05
Publication date: 1999-05-20
Anticipated expiration: 2000-05-07
Also published as: AU1016999A

Abstract

A cDNA clone for endo-β-mannanase which is prepared from either a fruit or the seed of the fruit is provided. The cDNA clone is used for the preparation of an antisense sequence which is useful for blocking the expression of endo-β-mannanase in various fruits. Since endo-β-mannanase is an enzyme involved with the ripening and softening of fruit, the antisense sequence is useful for delaying fruit ripening and resultant fruit softening. The invention also involves the use of promoters to over express the endo-β-mannanase gene in fruits, in order to expedite the rate of fruit ripening. A method is therefore provided for delaying the ripening of a fruit comprising the step of obtaining a cDNA clone for endo-β-mannanase produced by the fruit and inhibiting the expression of endo-β-mannanase in the fruit. A method is also provided for expediting the ripening of a fruit by comprising the step of inserting a promoter for the endo-β-mannanase gene into the genome of a fruit.

Description

cDNA Clone of Endo-β-Mannanase from Plant Tissues

Field of the Invention

The invention relates to the preparation of a cDNA clone for endo-β-mannanase from plant tissues for controlling the ripening of fruits.

Background of the Invention

Tomato fruit production on an annual basis worldwide is now second only to citrus in terms to weight of crop. Approx. 60-65% of total crop weight (approx. 50 million tonnes) is used for processing (FAO Production Yearbook). Cultivars suitable for mechanical harvesting for processing and fresh-market production require uniform setting and maturation of fruits; 75-90% of the fruits must be mature at the time of harvest (Lutyanenko, Genet. Sel. 55, 82-92, 1975). Slow maturing and softening of matured fruit can ensure their retention on the plants for 20-25 days without loss of marketing quality. Tomato ripening is a complex process involving a number of chemical and physical changes which convert the fruit from a relatively inedible state to one of optimal quality for processing and consumption. The events are highly synchronised following the onset of ripening, which is regarded as the initiation of a respiratory rise and associated ethylene production (Tigchelaar et al, HortSci. 13, 508-13, 1978). Measurable changes in ripening include changes in carotenoid and chlorophyll biosynthesis (with visible symptoms of colour change from green to yellow/orange to red), and decline in firmness of the fruit. Changes associated with ripening have a significant impact on the economics of fresh-market tomato cultivation, for example, and there is frequently an interval of two weeks between harvest of the fruit and availability to the consumer. This necessitates picking of the fruit at the mature-green stage and, thereafter, ripening occurs in transit. Shipments usually contain tomatoes at various stages of maturity, so fruit must be resorted and repacked during distribution with concomitant economic losses. Tomato cultivars with a controlled or increased shelf life could be harvested later in the ripening process. Shipments of such fruits would be of a higher quality, be more uniform at maturity, and spoil more slowly than those in mixed shipments. A large number of induced tomato mutants exist which exhibit decreased softening and prolonged shelf-lives (e.g. Nr, rin and nor) (Tigchelaar et al, Tomato Genet. Coop. 23, 33-36, 1973; Khudairi, Amer. Sci. 60, 696-707, 1972), but they are pleiotropic mutants that exhibit undesirable traits associated with fruit coloration and abnormal taste.

Our understanding of the ripening process has improved with the advent of advanced analytical, biochemical and molecular techniques, and one enzyme that was determined to play a role in fruit maturation is polygalacturonase (PG). This was the first fruit-ripening-related enzyme for which cDNA was isolated and sequenced (Grierson et al, Nuc. Acids Res. 14, 8595-603, 1986). The enzyme exhibits three major isoforms in ripening tomato fruit (PG1, PG2a and PG2b) which arise from a single gene (Bird et al, Plant Mol. Biol ϋ, 651-62, 1988). Activity of PG1 has been associated with the rate at which pectins becomes solubilized in the fruit (Delia Penna et al, Plant Physiol. 94, 1882- 6, 1990), although there is doubt that it is the primary or only determinant of softening (Hobson and Grierson, In: Biochemistry of Fruit Ripening, Seymour et al, eds, Chapman and Hall, pp. 405-42, 1993). Reduction in PG activity by anti-sense technology in the cvs. Ailsa Craig, and UC82B prevents pectin depolymerization and alters the firmness of the fruit (Kramer et al, Postharvest Biol. Technol. 1, 241-55, 1992); low PG tomatoes are also less susceptible to mechanical damage and cracking. The anti-sense experiments to remove PG activity were upgraded to a commercial scale by Calgene, and the FlavrSavr tomato was briefly marketed. This transgenic plant and product met all required EPA and USDA regulations; nutritional status (RDAs), taste, horticultural and developmental traits, and potential toxins (solanine and tomatine) were unchanged compared to non-transformed tomato varieties (Redenbaugh et al, Safety Assessment of Genetically Engineered Fruits and Vegetables. A Case Study of the FlavrSavr tomato, CRC Press, 1992). FlavrSavr tomatoes have now been withdrawn from the market, however, officially because the transformed cultivar does not meet required production and marketing requirements. An increased understanding of the ripening process has prompted a re-evaluation of the role that PG plays in fruit softening. The fruit cell wall consists of protein and three major polysaccharide components, pectin, hemicellulose and cellulose (O'Neill et al, In: Methods in Plant Biochemistry 2, 415-41, 1990). The hemicellulose polymers are comprised of xyloglucans, glucomannans and galactoglucomannans, which are covalently linked to pectin, and hydrogen-bonded to cellulose microfibrils. In growing plant tissues it is well accepted that it is the breaking of the hemicellulose components between the cellulose microfibrils that causes wall softening and turgor-driven cell expansion, and not changes to either the cellulosic or pectic components. Thus, targeting the hemicellulose component, of which mannans are a major constituent, as an agent of wall softening in fruits is a reasonable and sensible alternative, and the enzyme endo-β- mannanase is the key enzyme in the hydrolysis of mannans, through random cleaving of βl- 4 links in the backbone structure.

Endo-β-mannanase is an important enzyme in the mobilization of cell wall reserves in seeds, and was first reported as a secreted enzyme in the endosperms of endospermic legumes (McCleary, Phytochem. 14, 1187-94, 1975). Recent studies have revealed it to be present in the seeds of more than twenty species of monocots, dicots and gymnosperms (and all cultivars or provenances thereof), usually in numerous isoforms (Dirk et al, Phytochem. 40, 1045-56, 1995). Very few studies have been conducted to determine if mannanase is present in fruit tissues, but it has been reported to be present in those of tomato (Pressey, Phytochem. 28, 3277-80, 1989), and recently confirmed

(Bewley, unpublished). Cell wall thickening is frequently associated with the deposition of one particular type of polysaccharide, including galactomannan-type hemicelluloses.

These contain a rigid mannose backbone with substituted galactose or galactose and glucose side chains, and they provide rigidity to the wall, in conjunction with cellulosic and other hemicellulosic components. The endosperms of some extremely hard seeds, e.g. date palm and ivory nut, are composed largely of galactomannan-containing cell walls (Bewley and Reid, In: Biochemistry of Storage Carbohydrates in Green Plants, Academic Press, N.Y., 1985 pp. 289-304). While the galatcomannan-containing cell walls of other seeds and of fruits are not as rigid, they are important in maintaining structural integrity, and their softening requires endo-β-mannanase.

Research progress on endo-β-mannanase has been relatively slow, and studies largely confined to its physiological activities. The impediments have been:

(a) The only assay for endo-β-mannanase was a time-consuming viscometric one, which severely limited the number of samples that could be tested.

(b) No endo-β-mannanase had been purified in sufficient quantities to allow for N-terminal sequencing, for while it is an extremely active enzyme, it is produced by seed tissues only in very small amounts.

(c) No cDNA clone was available for the plant enzyme. The first observation of endo-β-mannanase in fruit was made by Pressey in his article entitled "Endo-β-mannanase in tomato fruit" (Phytochemistry, 28, 1989:3277- 3280). Pressey detected very low levels of endo-β-mannanase in extracts prepared from tomato fruit tissue. The levels of endo-β-mannanase detected in the tomato fruit were so low that it could not be concluded from the study that endo-β-mannanase plays a role in fruit ripening. In fact, this research was never followed up. Pressey did observe an increase in the level of endo-β-mannanase during ripening and conjected that this may suggest a role for the enzyme in the ripening process. However, the miniscule amounts of endo-β-mannanase detected by Pressey led to the conclusion that the role of endo-β- mannanase in the ripening process remains unknown and that further studies were needed. It was not thought that these further studies were worthwhile and as a result no one in the field followed up Pressey' s work regarding the role of endo-β-mannanase in fruit ripening.

There is a need for the isolation of an enzyme which is effective for catalysing the ripening of fruit. There is a further need for the development of a cDNA clone for this enzyme so that expression of the enzyme can be suppressed from known techniques such as anti-sense technology. There is also a need to overexpress this enzyme in certain situations through the use of promoters in order to expedite the rate of fruit ripening. Summary of the Invention

The present invention is a cDNA clone for endo-β-mannanase. The invention also involves a method for controlling the ripening of fruit using a cDNA clone for endo- β-mannanase.

The invention permits for the preparation of an antisense sequence from the cDNA clone which was obtained from either a fruit such as tomatoes, melons, peaches, oranges, cucumbers and nectarines or a seed such as a tomato seed. The antisense sequence can be inserted into the genome of a fruit for blocking or significantly inhibiting the expression of endo-β-mannanase. Since the fruit endo-β-mannanases are at least 60% homologous to the seed endo-β-mannanases, therefore the seed endo-β-mannanases can also be used to prepare antisense sequences useful in fruits. Also promoters for endo-β- mannanase can be inserted into the genome of a fruit in a manner known in the art for over expressing the endo-β-mannanase gene. According to one aspect of the invention, a cDNA clone for endo-β-mannanase is provided. According to another aspect of the invention, a cDNA clone that is at least 60% homologous to a cDNA clone for endo-β-mannanase is provided. Homology refers to the fact that this cDNA clone has a nucleic acid sequence that is at least a 60% match in primary structure to that of the cDNA clone for endo-β-mannanase. According to another aspect of the invention, an expression system comprising a cDNA clone for endo- β-mannanase and a promoter for expressing same is provided. According to another aspect of the invention, an antisense sequence for endo-β-mannanase for inhibiting the expression of endo-β-mannanase is provided. According to yet another aspect of the invention, a method for delaying the ripening of fruit is provided comprising the step of obtaining a cDNA clone for endo-β-mannanase produced by a fruit and inhibiting the expression of endo-β-mannanase in the fruit. According to another aspect of the invention, a method for expediting the ripening of a fruit is provided comprising the steps of inserting a promoter for endo -β-mannanase into the genome of a fruit and overexpressing the endo-β-mannanase in the fruit. These and other aspects are decribed in greater detail hereinbelow. . Brief Description of the Drawings

Figure 1 is a complete nucleotide sequence and deduced amino acid sequence of the cDNA clone pDB-MAN. Amino acid numbering begins at the NH₂-terminus of the mature enzyme, which is indicated by an arrow. The 65 NH₂-terminal amino acid residues from the purified protein are underlined and a possible N-glycosylation site in this region is underlined. The putative catalytic acid (Glu-148) and nucleophile (Glu- 265) residues are also underlined. The stop codon is marked by an asterisk and a putative polyadenylation signal is underlined.

Figure 2 is an amino acid sequence alignment of the Trichoderma (trr), Aspergillus (asn) and tomato (man) (1— >4)-β-mannanase. The alignment was performed using the PRETTYBOX program. Amino acid numbering begins at the translation start methionine residues. Asterisks indicate the putative catalytic residues. Figure 3 is a Genomic Southern blot analysis of DNA isolated from tomato seedlings (cv. Dynamo FI) cut with.YZ.aI (Xb), Hindlll (H), and^TzoI (Xh) and probed with the complete tomato cDNA.

Figure 4 is a Northern blot analysis of total RNA isolated from various tomato tissues. Lanes: 1 WS, whole seeds, 2 SR, seedling roots; 3 SH, seedling hypocotyls; 4 SC, seedling cotyledons, 5 ML, mature leaves; 6 MR, mature roots; 7 E, 72-h endosperm; 8 C, 72-h cotyledons.

Figure 5 is a table setting out the activities of endo-β-mannanase in various different fruits.

Figure 6 is a plot of stage of ripening versus endo-β-mannanase activity showing an endo-β-mannanase time course for ripening tomato.

Figure 7 is a Western blot showing the homology of endo-β-mannanase obtained from tomato seed to the endo-β-mannanases of various different fruits. Detailed Description of the Invention

It is an aspect of the present invention to provide a c-DNA clone for endo-β- mannanase. In specific embodiments of the invention, the c-DNA clone is for endo-β- mannanase isolated from a fruit. More specifically, the c-DNA clone is for endo-β- mannanase isolated from a fruit selected from the group consisting of tomatoes, melons, peaches, oranges, cucumbers and nectarines. Most specifically, there is provided a c- DNA clone for endo-β-mannanase isolated from a tomato seed. In another embodiment of the invention, the c-DNA clone is for endo-β-mannanase isolated from a fruit and this c-DNA clone is at least 60% homologous to a c-DNA clone for endo-β-mannanase obtained from a tomato seed.

Oligonucleotides corresponding to NH₂-terminal amino acid sequences of a purified tomato (l→4)-β-mannanase (Nonogaki et al. Physiol. Plant, 94, 1995:328-334) were used to screen a cDNA library which had been generated from poly(A)+ RNA of 3-d-germinated tomato seeds. One near full-length cDNA clone of 1307 bp was isolated and its complete nucleotide sequence was determined (Fig. 1). The cDNA was shown to encode a tomato (1— >4)-β-mannanase because the amino acid sequence deduced from a region near its 5' end (nucleotides 88-282) is identical with the sequence of the 65 NH₂- amino acids determined directly from the purified enzyme. A region encoding a putative signal peptide of 23 amino acid residues is observed

5' to the codon which specifies the NH₂-terminal Asn residue (Fig. 1). The putative signal peptide is typical of eukaryotic signal peptides (Watson, Nucleic Acids Res. 12, 1984:5145-5164). The nucleotide sequence indicates that the mature enzyme has 346 amino acid residues and a calculated MR of 38 950. This can be compared with a value of 38 000 reported for the (l-»4)-β-mannanase isoform MI purified from tomato endosperm (Nonogaki and Morohashi. Plant Physiol. 110, 1996:555-559). The pi calculated from the primary structure of the enzyme is 5.3; isoelectric focussing of endosperm extracts from germinated tomato seeds revealed the presence of several isoforms with pi values in the range 5.2-5.8 (Toorop et at. Planta, 200, 1996:153-158; Voigt and Bewley, Planta, 200, 1996:72-77).

Alignment of the tomato enzyme with fungal (1— »4)-β-mannanases shows overall positional identities of 30% with the Aspergillus aculeatus enzyme and 28% with the Trichoderma reesei (1— »4)-β-mannanase (Fig. 2). Short, conserved blocks of two to four amino acid residues are dispersed along the polypeptide. Conserved glutamic acid (Glu) residues at positions 144, 148, 185 and 265 (Figs 1, 2) are candidates for catalytic residues and their possible participation in catalysis can now be investigated using chemical modifications and irreversible inhibitors (cf. Chen et al. J. Biol. Chem. 263, 1993 : 13318- 13326). The catalytic domains of several hundred glycosyl hydrolases have been classified into distinct families, based on similarities in amino acid sequences and hydrophobic cluster analyses (Henrissat and Bairoch, Biochom, 293, 1993:781-789). The tomato (1— »4)-β-mannanase sequence determined here indicates that the enzyme is a member of family 5. According to this classification the catalytic acid of the tomato (l-»4)-β-mannanase would be expected to be Glu-148, the catalytic nucleophile would be Glu-265, and the anomeric configuration of mannosyl residues released during hydrolysis would be retained (Henrissat and Bairoch, Biochom, 293, 1993:781-789). The crystal structures of two family 5 (l- 4)-β-glucan endohydrolases from Clostridium spp. have been solved; the enzymes display a (β/α)s fold. Highly conserved amino acid residues involved in substrate binding and catalysis have been identified (Ducros et al., Structure, 3, 1995:939-949; Dominguez et al. J. Mol. Biol., 257, 1996:1042-1051). Six of these conserved amino acid residues can also be identified in similar positions in the tomato (l-»4)-β-mannanase (Arg-58, His-108, Asn-147, His-223, Tyr-225 and Trp-303; Fig. 1).

Although no other sequences of plant (l-»4)-β-mannanases have been published or appear in the databases, the primary structures of several microbial (1— >4)-β- mannanases have been deduced from the nucleotide sequences of cloned DNA fragments. As mentioned above, alignment of the plant enzyme sequence with sequences from fungal (l-»4)-β-mannanases reveals significant sequence identity (Fig. 2) but inclusion of bacterial sequences greatly reduces the number of identical residues (data not shown). It can be concluded that the bacterial enzymes exhibit little, if any, sequence similarity with the eukaryotic enzymes, and presumably evolved their (1— »4)-β-mannanase specificity through convergent evolution from unrelated ancestral proteins. Similar conclusions have been reached for the evolution of bacterial and plant (1— »3, 1— 4)-β- glucanases (Hoj and Fincher, Plant J., 7, 1995:367-379).

The alignments presented in Fig. 2 indicate that, in comparison with the tomato enzyme, fungal (1— >4)-β-mannanases are sometimes extended at their COOH-termini. These COOH-terminal extensions presumably correspond to the protein docking domains or additional catalytic domains that have been reported in many microbial (l→4)-β-mannanase (Gibbs et al., Appl. Environ. Microbiol. 58, 1992:3864-3867; Fanutti et al., J. Biol. Chem. 270, 1995:29312-29322). The tomato (l→4)-β-mannanase does not appear to have multiple domains of this type (Fig. 2).

Southern hybridization analysis of genomic DNA cut with several restriction enzymes indicated that tomato (l→4)-β-mannanases are encoded by a family of approximately four genes (Fig. 3). This observation must be reconciled with the multiple (1— »4)-β-mannanase isoforms that have been separated from extracts of germinated tomato seed by isoelectric focussing, where up to 15 active (1— »4)-β-mannanase bands are visible (Fig. 4) (Toorop et at. Planta, 200, 1996:153-158; Voigt and Bewley, Planta, 200, 1996:72-77). The obvious conclusion is that many of the isoforms present on isoelectric focussing gels are not true genetic isoenzymes, but are generated from a relatively small number of (1— 4)- β-mannanase by post-translational modifications. Differential glycosylation patterns and the exposure of the enzymes to amino- and carboxy-peptidases in the germinated seed could be responsive for these effects. Although the washing stringency used in these Southern analyses was chosen to ensure that related (1 - 4)-β-mannanase genes were detected, it is also possible that other more divergent (l- 4)-β-mannanase genes were not detected by the probe under the hybridization conditions used here.

The (1— »4)-β-mannanase cDNA was used to investigate expression sites of the genes in several tissues, as determined by mRNA transcripts detected in Northern hybridization analyses. (1— »4)-β-mannanase mRNA transcripts were only defected in the endosperm of seed 72 h after germination (Fig. 4). This is consistent with the detection of the enzyme itself in this tissue, where its synthesis is inducible by gibberellin (Toorop et at. Planta, 200, 1996:153-158; Nonogaki and Morohashi, Plant Physiol. 110, 1996:555- 559; Still and Bradford, Plant Physiol. 113, 1997:21-29; Still et al. Plant Physiol. 113, 1997:13-20). However, the presence of at least 4 (1— 4)-β-mannanase genes in tomato (Fig. 3) indicates that detailed studies on expression patterns and their regulation by phytohormones or other developmental factors will only be possible when gene-specific probes become available. The cDNA described here (Fig. 1) can now be used to isolate other tomato (l-»4)-β-mannanase cDNAs or genes, for the generation of gene-specific DNA fragments.

A major difficulty in studying the function and developmental regulation of (1— »4)-β-mannanases in germinated tomato seeds is the very small size of the seeds themselves and the associated difficulties in isolating sufficient mRNA or purified enzyme for analysis. The availability of cDNA probes presents opportunities to monitor (l→4)-β-mannanase gene expression in small seeds by in-situ hybridization histochemistry (McFadden et al. Planta, 173, 1988:500-508). It should also allow (1→4)- β-mannanase genes from other plant species to be isolated for similar developmental studies. Finally, expression of the cDNA in heterologous systems should allow sufficient enzyme to be synthesized for thorough chemical and kinetic analyses, and for detailed examination of substrate specificities, catalytic mechanisms and action patterns of the enzyme. It is therefore another aspect of the present invention to provide a c-DNA clone for use in the monitoring of (l→4)-β-mannanase gene expression, isolation of (l→4)-β- mannanase genes from other plant species and synthesis of (1— »4)-β-mannanase through expression in heterologous systems.

A cDNA for endo-β-mannanase is isolated from a fruit including tomatoes, peaches, oranges, nectarines and melons or others according to the same procedure as is used for obtaining the cDNA from the tomato seed with the necessary modifications as would be apparent to a person skilled in the art. Since mannans are a major cell wall component in ripening tomato^* fruits, and since the enzyme, endo-β-mannanase, which degrades them is present at the time of ripening, the activity of latter can be eliminated using anti-sense technology, and assessed as to the extent to which fruit firmness is retained. Ripening fruits of selected cultivars are assessed for endo-β-mannanase activity during fruit development and ripening. Different parts of the fruit are analysed; including the locular tissue, seed, seed sheath, skin, and outer pericarp tissue immediately beneath the skin. Preliminary studies on a supermarket-bought bilocular and multilocular cultivar, and one 'plum' (processing) cultivar have shown that endo-β-mannanase activity increases following the breaker stage, is prominent during the green to orange stage, and declines as the fruit becomes over-ripe and deep red. A single isoform of the enzyme of approx. pI9 is present. Further experiments on a defined cultivar are carried out to determine the precise timing of enzyme production and activity during ripening, and which tissues produce it. Preliminary data strongly suggest that it is produced only in the region beneath the skin, the region most important in the retention of fruit firmness. The assay for the enzyme is colorimetric, and involves the incorporation of locust bean gum galactomannan into Phytagel on an assay plate. Enzyme extracts are placed in wells in the plate, incubated at 37°C and the extent of substrate breakdown assessed using Congo Red dye (Downie et al, Phytochem. 36, 829-35, 1994). The area of clearing of substrate around the wells can be accurately and quantitatively determined using Aspergillus enzyme standards. Enzyme production and visible changes to the fruit are correlated with changes in fruit firmness and skin resistance. Standard tests include force-deformation using a dynamometer equipped with a flat disc, resistance to puncture using a table dynamometer with a sharp point, and damage resulting from a free-fall, usually from 70 cm (Altisent et al, Symposium on Production of Tomatoes for Processing, ISHS, pp 181-96, 1979). Various points on the surface of the fruit are tested. The timing and location of the expression of the gene for endo-β-mannanase is determined in the developing fruit since this controls the appearance of the enzyme. Total RNA, or messenger RNA following oligo dT-column chromatography, is separated on an agarose gel and detected by northern hybridization, using a ³²P-labelled tomato seed endo-β- mannanase cDNA clone as the probe. In this way, necessary correlations are drawn between gene transcription, enzyme production and fruit ripening characteristics. Mutants defective in their ability to ripen are also used to correlate endo-β-mannanase activity with fruit softening, to determine if the rin, nor and ale genes, for example, directly or pleiotropically affect enzyme production. It is another aspect of the present invention to generate an anti-sense construct for endo-β-mannanase for inhibiting the expression of endo-β-mannanase. Specifically, the antisense construct is for endo-β-mannanase obtained from a fruit, more specifically, a tomato and even more specifically, a tomato seed.

In another aspect of the present invention, there is provided a method to delay the ripening of a fruit. In an embodiment of the invention, this method comprises the steps of obtai ing a c-DNA clone for endo-β-mannanase produced by the fruit and inhibiting the expression of the endo-β-mannanase in the fruit. In a specific embodiment of the invention, the method further includes the steps of preparing an antisense sequence from the c-DNA clone and inserting the antisense sequence into the genome of the fruit. In more specific embodiments the fruit is selected from the group including tomatoes, melons, peaches, oranges, cucumbers and nectarines. In the most specific embodiment of the invention, the fruit is tomato.

In another of its aspects, the present invention also provides a method for expediting the ripening of a fruit. In specific embodiments, this method comprising the steps of obtaining a cDNA clone for endo-β-mannanase and inserting a promoter for endo-β-mannanase into the genome of the fruit in order to over express the gene for endo-β-mannanase . In more specific embodiments, the fruit is selected from the group including tomatoes, melons, peaches, oranges, cucumbers and nectarines. In the most specific embodiment of the invention, the fruit is tomato. The- present invention also embodies the use of a cDNA clone for endo-β- mannanase of a fruit for controlling the rate of ripening of the fruit. In a specific embodiment, this use is directed to a fruit selected from the group comprising tomatoes, melons, peaches, oranges, cucumbers and nectarines. In another specific embodiment, this use is specifically directed to delaying the ripening of a fruit. It is also another specific embodiment to use a cDNA clone for endo-β-mannanase of a fruit to expedite the ripening of the fruit. Finally, the present invention also embodies the use of an antisense sequence for endo-β-mannanase for delaying the ripening of a fruit.

Tomato is a crop plant with a relatively small DNA content per haploid genome, well-developed genetics and favourable tissue-culture characteristics. Efficient transformation procedures permit the introduction of foreign genes into explants and protoplasts which can be regenerated into transgenic, fertile plants (Hille et al, In: Genetic Improvement of Tomato, Monograph on Theoretical and Applied Genetics H, 283-291, 1991). Several protocols for transforming tomato have been published, including:

McCormick et al, Plant Cell Rep. 5, 81-84, 1986; McCormick, Plant Tissue Culture Manual B6, 1-9 1991; Fillati et al, Biotech. 5, 726-730 1987. These procedures use leaf disks as explants, like the original leaf disk transformation protocol that was described for plants (Horsch et al, Science 227, 1229-1231, 1985), or cotyledon explants. The typical vector system that is used for transformation is an Agrobacterium tumefaciens strain with a disarmed Ti plasmid and an additional plasmid carrying the gene of interest and a selectable marker gene (like NOS-NPTII-NOS for kanamycin resistance) within the T-DNA region (McKormick et al, 1986, Plant Cell Rep. 5:81). The gene of interest is introduced into the plant material by cocultivating the explant and the Agrobacterium for a few days, killing the bacteria with an antibiotic and selecting shoots which regenerate from the explant on media containing the selection agent (kanamycin). DNA samples from shoots that are resistant to the selection agent are then tested to determine whether they also contain the gene of interest.

Tomato fruits from the transformed plants are assessed for their ability to produce endo-β-mannanase, and to transcribe the mRNA for this enzyme. Comparisons are made with the time course and quantitative expression determined for the fruits from untransformed plants. Tests for firmness and skin resistance are likewise conducted, and visible changes in fruit characteristics noted. Most facets of fruit development and ripening are unaffected, but firmness from the breaker stage is retained and endo-β- mannanase activity is eliminated or severely curtailed.

Fruits on transgenic tomato plants are produced that retain their firmness for a longer period than non-transformed plants, and thus are more uniformally developed as far as ripening, harvesting, longevity and storability are concerned. This produces benefits related to harvesting of market-ready and processing tomatoes, in that a wider window of time is available, there is a reduction in the number of times that marketed fruits will have to be handled and sorted, and a reduction in losses due to deterioration.

Experimental Examples

Experiment 1 : Germination and seedling growth.

Seeds or tomato (Lycopersicon esculentum Mill. cv. Dynamo FI; Sandoz seeds, supplied by B. & H. Stoeff, Virginia, S. Australia) were washed in sterile distilled water to remove the Thiram seed-coat treatment and placed on two layers of Whitman No. 1 filter paper in 9-cm Petri dishes on sterile water. When the radicles and hypocotyls had grown to the appropriate length under ambient laboratory conditions (approx. 23 °C) in darkness, the seedlings were harvested, surface-dried by blotting with a paper towel, frozen in liquid N₂, and stored at -8O°C.

Experiment 2: Analysis of NH₂-terminal sequence.

Tomato seed (1— »4)-β-mannanases form MI was purified as described by Nonogaki et al. 94, 328-334, 1995. The NH₂-terminal amino acid sequence was determined in a Hewlett-Packard G1OO5A protein sequencer (Hewlett Packard Company, Palo Alto, Calif., USA) using the Hewlett-Packard 3.0 sequencing routine, which is based on Edman degradation chemistry. Experiment 3: Isolation of poly(A)+RN A.

Approximately lg dry weight of mature tomato seeds was germinated and total RNA was extracted from young seedlings when the radicles had grown 5-8mm (3-4d from imbibition). The method of Prescott and Martin (Plant Mol. Biol. Rep., 4, 219-224, 1987) was used to isolate poly(A)+RNA. Seedlings were ground to a powder in liquid N₂ and suspended in a total of 7.5 ml extraction buffer (50 mM Tris-HCl buffer, pH 9, containing 150 mM LiCl, 5 mM EDTA, 5% w/v SDS) before phenol: chloroform extraction. The RNA was precipitated overnight at 4°C with 2 M LiCl, and after dissolving in water the poly(A)+RNA fraction was obtained by oligo(dT)-cellulose chromatography (Aviv and Leder, Proc. Natl. Acad. Sci. 69, 1408-1412, 1972).

Experiment 4: Isolation of cDNA clones.

A cDNA library was prepared from 5 μg poly(A)+RNA of germinated tomato seeds using the λZAP cDNA synthesis and cloning kit (Uni-ZAP XR; Stratagene, La Jolla. Calif, USA) according to the manufacturer's instructions. The library was screened by hybridization of nitrocellulose filter plaque replicas with three degenerate oligonucleotides which were designed on the basis of NH₂-terminal amino acid sequence data from the purified tomato seed l→4)-β-mannanases and end-labeled with [γ^PjATP.

The cDNA inserts of positive clones were rescued into the pBluescript SK(+) phagemid and both strands were sequenced using the dideoxynucleotide chain termination procedure (Sanger et al. Proc. Natl. Acad. Sci. 74, 5463-5467, 1977). Computer analyses were effected with the University of Wisconsin Genetics Computer Group software (Devereux et al. Nucleic Acids Res., 12, 287-395, 1984) in the ANGIS suite of programs at the University of Sydney.

Experiment 5: Extraction of RNA and Northern blot analysis. Total RNA was prepared using the procedure of Prescott and Martin (Plant Mol. Biol. Rep. 4, 219-224, 1987) described above from 150 intact seedlings that had been germinated for 72 h (radicle lengths 5-10 mm), from 120 endosperms and cotyledons dissected from 72-h-germinated seeds, from 150 roots (2-2.5 cm in length), 150 hypocotyls (2.5-3 cm in length) and 150 cotyledons (unexpanded and still confined within the seed coat) from 7-d-germinated seedlings, and from 1.5 g leaves and 1.8 g roots of mature, greenhouse-grown tomato plants. Samples containing, where possible, 5 μg RNA were separated on a 1.2% agarose gel containing 2.2 M formaldehyde, transferred onto a Hybond-N+ (Amersham) nylon membrane, and probed with a [³²P]- labelled. 1.1 -kb EcoRI -Xhol fragment of the (l→ )-β-mannanases cDNA clone (Sambrook et al. Molecular Cloning: A Laboratory Manual, 2^nd Εd., Cold Spring Harbour Laboratory Press, Cold Spring Harbour, N.Y., 1989). Following hybridization the membrane was washed progressively to 0.5 x saline sodium citrate (1 x SSC = 0.15 M NaCl, 0.015 M Na₃citrate pH 7) and 0.1% SDS at 65°C. Experiment 6: Isolation of DNA and Southern blot analysis.

Approximately 2 g fresh weight of young tomato leaves, cultivar Dynamo FI, were extracted in a 30-ml volume of 100 mM Tris-HCl buffer (pH 8.0) containing 50 mM EDTA (pH 8.0) and 10 mM 2-mercaptoethanol. Sodium dodecyl sulfate was added to 1% and after heating to 65°C for 10 min, potassium acetate was added to 1 M. Following chilling on ice and centrifugation , the supernatant was filtered through Miracloth

(Calbiochem, La Jolla, Calif, USA), precipitated with 0.7 vol. isopropanol and the pellet redissolved in 50 mM Tris-HCl buffer (pH 8.0) containing 10 mM EDTA (TE buffer).

Following reprecipitation with wopropanol and washing with 70% ethanol, the pellet was redissolved in TE buffer and extracted twice with phenol and chloroform, precipitated with 2.5 vol. ethanol and re-dissolved in TE buffer. The DNA was digested with restriction endonucleases Xbal, Hindlll and Xhol and separated on a 1% agarose gel before transfer to a Hybond-N+ nylon membrane (Amersham). The membrane was probed with a [³²P] -labelled full-length fragment of the tomato l- 4)-β-mannanases cDNA and, following hybridization, was washed progressively to 0.5 x SSC+0.1% SDS at 55°C.

Summary of Results from Experiments 1-6:

Isolation and characterisation of cDNAs. Based on the sequence of the 65 NH₂- terminal amino acids of the purified tomato (l→4)-β-mannanase, three redundant oligonucleotides of 21-24 nucleotides in length were synthesised. Approximately 5 x, 10⁵ clones were screened with a mixture of the 3 oligonucleotides and 19 positive clones were isolated. The cDNA inserts ranged in size from 500 bp to 1.3 kb; the clone containing the largest cDNA was rescued into the pBluescript SK(+) phagemid and was designated pDB-MAN The complete nucleotide sequence of the 1307-bp cDNA, together with the amino acid sequence deduced from it, are shown in Fig. 1. The cDNA has an open reading frame which extends from nucleotides 19 to 1125. The amino acid sequence deduced from nucleotides 88 to 282 corresponds exactly with that of the 65 NH2 -terminal amino acids determined directly from the purified tomato (1— >4)-β-mannanase isoenzyme MI (Fig. 1). We conclude, therefore, that the cDNA encodes the tomato (l-»4)-β-mannanase.

In the region of the cDNA from nucleotide 88 to the first in-frame stop codon at nucleotide 1126, there is an overall (G+C) content of 37.6%, which reflects the relatively balanced codon usage in the tomato (1— »4)-β-mannanase gene (Fig. 1 and data not shown). Immediately 5' to the codon for the NH₂-terminal asparagine (Asn) residue is a region encoding a putative signal peptide of 23 amino acid residues, together with a short

5' untranslated region (Fig. 1). At the 3' end of the coding region of the mature enzyme are two consecutive TGA stop codons, which form part of a 182-bp 3' translated region.

A 21-bp poly(adenylic acid) tail is evident, and a putative AATAAA polyadenylation signal begins at nucleotide 1255 (Fig. 1).

Properties of the encoded enzyme. The region of the cDNA from the codon for the NH₂-terminal Asn residue at nucleotide 88 to the stop codon at nucleotide 1126 encodes the mature enzyme. The enzyme has 346 amino acid residues, a calculated Mr, of 38 950 and an estimated pi of 5.3. A single potential N-glycosylation site (Asn-Gly- Ser) begins at amino acid residue 15 in the mature enzyme sequence. No myristylation or phosphorylation consensus sites could be detected.

Overall amino acid sequence identities of 30% and 28% are observed between the tomato (1— 4)-β-mannanase and those from Aspergillus aculeatus (Christgau et al. Biochem. Mol. Biol. Internat. 33, 917-975, 1994) and Trichoderma reeseii (unpublished database entry, accession number L25310), respectively. The alignment ofthese three sequences is shown in Fig. 2, where short blocks of identical residues are seen to be distributed along the length of the sequences.

Southern hybridization analysis. A Southern blot of tomato genomic DNA cut with restriction enzymes and probed with the complete tomato (1— >4)- β-mannanase cDNA revealed the presence of multiple hybridizing DNA fragments (Fig. 3). Based on these observations, we conclude that tomato mannanases are encoded by a gene family of about four genes and that the numerous hybridizing bands in the Xbal digest (Fig. 3) result from relatively abundant Xbal sites in introns in the genes. Northern hybridization analysis. Total RNA was isolated from tomato seeds germinated for 3 d, and from cotyledons and endosperm dissected from those germinated seeds. Total RNA was also extracted from tissues of 7-d-old seedlings, including roots, hypocotyls and cotyledons. These RNA samples were subjected to Northern hybridization analysis, together with RNA from the roots and leaves of mature plants, and the cDNA encoded tomato (l-»4)-β-mannanase isoform MI was used as a probe (Fig. 4). Ethidium bromide staining of the agarose gel before transfer of the RNA showed approximately equal loadings of RNA from the 72-h-germinated seed, 72-h dissected cotyledons, 7-d seedling root and hypocotyl samples, and from 7-d seedling cotyledon sample. However, only approximately half the amount, or less, was loaded from the leaf and root samples, and ethidium bromide-stained RNA was undetectable in the lane in which the 72-h endosperm sample was loaded. Yields of RNA from these last samples were extremely low; it was difficult to dissect out sufficient tissue to scale up the isolation and it was therefore not possible to prepare microgram quantities of RNA. Furthermore, comparisons of amounts of mRNA based on differences in hybridization intensity in Fig. 4 must be viewed with caution.

Nevertheless, it was clear that mRNA transcripts encoding the (1— 4)-β- mannanase were present in extracts of the endosperm of seeds 72h after germination; no (1— 4)- β-mannanase mRNA was detected in cotyledons at this stage. Furthermore, the (l→4)-β-mannanase could not be detected in tissues of 7-d-old seedlings, nor in vegetative tissues of mature plants (Fig. 4).

Experiment 7: Enzyme Assay.

The activity of endo-β-mannanase in various fruit tissues is shown in Figure 5. This data was obtained by following the procedure as set out below: Endo-β-mannanase activity was assayed using a modified gel diffusion method (Downie et al., Phytochemistry, 40, 1045-1056; Toorop et al., Planta, 200, 153-158, 1996; Still et al., Plant Physiol., 113, 21-29, 1997) based on diffusion of the enzyme through an agarose gel-Locust bean galactomannan substrate matrix from a 2 mm diameter central well. The matrix was prepared by preheating 0.1 M citrate /0.2M Na₂HPO₂ buffer (pH 5.0) just to boiling in a microwave or a heating plate. Locust bean galactomannan substrate (LBG) (0.1% w/v) was added to the incubation buffer then stirred continuously for 2 hours on a heat plate at 80°C. To aid in dissolving the gum, the solution was continuously stirred overnight for 12 h at room temperature then finally centrifuged at 4000g for 15 min at 10°C. The supernatant was collected in aliquots of 50 mL. To make a uniform gel layer, a vertical cassette consisting of 0.5 mm U-frame spacer, and a cover glass plate (Pharmacia Biotech, Uppsala, Sweden) was prepared. The spacer was made hydrophobic (to allow easier removal of the wet gel layer on) by spreading a few mL of repel silane (Pharmacia Biotech) over the inner face with a tissue under the fume hood. The chloride ions which result from the coating with repel silane was rinsed off with double distilled H₂0. A few mL of H₂0 was poured on the cover glass plate and the agarose gel bond support film (Pharmacia Biotech) was placed on it with the hydrophobic side down. This facilitates filling the cassette later on. The gel bond film was firmly pressed against the glass plate with a roller (Pharmacia Biotech), with the film overlapping the length of the glass pate by about 2 mm. The glass plate, with the gel bond film attached, was placed on the U-frame spacer and the cassette was firmly clamped (FLEXI CLAMP, Pharmacia) together. The cassette and a 10 mL pipette was prewarmed to 65°C in a ventilated oven (MEMMERT, Schwabach, Germany).

High EEO, Type III-A agarose (Sigma Chemicals) was added to the LBG substrate solution (0.8% w/v) while still cold. This agarose-substrate mixture was heated to boiling in a microwave oven (at the lowest setting) until the agarose was^' completely dissolved (rapid stirring damages the mechanical properties of agarose, which is why heating in the microwave oven is preferable). The hot mixture was cooled down for 15 mins to about 65 °C in the ventilated oven then quickly dispensed into the prewarmed cassette. The cassette was let stand for about 3 hours at 15°C to allow the gel to solidify before use. Storing the gel in a moist environment at 15°C for at least 24 h was observed to improve the property of the gel and could be safely stored for up to one week before use (the final agarose gel structure forms after a few days in storage). On the other hand, cracks were observed around punched holes on gels stored at 4°C when exposed to room temperature. Fifty-five 2 mm wells were punched in each gel with a plastic pipette tip and removing the plugs with fine forceps before pipetting 2 μl samples of enzyme extract in the wells. Removing the plugs by suction was observed to cause leakage under the wells during incubation of the extract and thus was abandoned. The gel was placed in a moistened plastic box (to prevent drying) covered with plastic film then incubated for 20 h at 25°C. Commercial endo-β-mannanase (Megazyme International Ltd., Wicklow, Ireland) purified from Aspergillus niger (specific activity, 38 Units/mg protein using carob galactomannan as substrate) was used as a standard. A standard curve was developed in the same gel in dilutions ranging from 0.418 to 0.001045 International Units. After incubation, the gel was prewashed for 30 mins in 0.1M citrate/0.2M

Na₂HPO₄ buffer (1:4.7 ratio, pH 7) at 30 rpm on an orbital shaker (Gesellschaft fur Labortechnik mbH, Burgwedel, Germany), then stained with 0.5% (w.v) Congo red dye (Sigma) in water solution for 30 min while continuously shaking. The gel was fixed in 80% (w/v) ethanol for 10 min, destained in 0.1 M citrate/0.2 M. Na₂HPO₄ buffer (1:4.7 ratio, pH 7) for 10 min, then transferred to 1 M NaCl solution while continuously shaking. The hydrolyzed zones around each well started turning red and the background completely dark blue within 10 to 25 minutes of transferring the gel to 1 M NaCl solution. Fixing in ethanol, and continuous shaking was observed to improve sharpness of the edges of the clearing zones around each well and falls in agreement with a previous observation by Still et al. (1997). Changing of NaCl solution several times while destaining also improved the contrast of the clearing zones from dark red to light red colour. An outer ring around the clearing zones previously observed to disappear with continuous destaining in NaCl solution within one hour was clearly preserved when the salt solution was changed three times in one hour (after 30, 45, and 60 mins). Changing to a fresh NaCl solution during destaining appears to improve clarity of the hydrolyzed zones by removing the surplus non-bound dye and cleaved dye binded polymeric galactomannan units which still adheres to the surface of the gel thus masking the edges of the clearing zones. NaCl apparently enhances the binding of the Congo red dye to the polysaccharides and turns the background colour darker. The clearing zones around the wells turn lighter since congo red binding is reduced by cleavage of polymeric units of galactomannan substrate by the enzyme. This suggests that the basis for the changing colour pattern along the gel is related to the dye content and the pH.

Endo-β-mannanase activity was calculated by measuring the diameter of the hydrolyzed zones manually with sliding callipers and comparing them with a standard curve of activity using Aspergillus endo-β-mannanase. This is a routine measurement which has been employed in previous investigations of endo-β-mannanase activity (Downie et al., Phytochemistry, 40, 1045-1056, 1994; Toorop et al., Planta, 200, 153- 158, 19

96; Still et al.,Plant Physiol. H3, 21-29, 1997; Still et al, Plant Physiol. H3, 13-20, 1997). A summary of endo-β-mannanase activity in various fruit tissue is provided in Figure 5. A graph showing the activity of endo-β-mannanase in the skin, outer pericarp and inner pericarp of the tomato over the time course ripening is provided in Figure 6.

Figure 7 shows the results of a western blot that was prepared by a method known in the art. The results show binding of an antibody for the endo-β-mannanase of the tomato seed. The results indicate that there exists at least a 60% homology in structure between the endo-β-mannanase of the tomato seed and the endo-β-mannanase of the fruits. Experiment 7: Transformation of Plants

For vector construction and Agrobacterium manipulation, the tomato mannanase antisense cDNA are inserted into the intermediate vector pKYLX attached to a double CaMV35S promoter, or the super-promoter now available under licence, between the borders of the t-DNA region and mobilized into the Agrobacterium strain LB A 4404 by triparental mating with an E. coli helper strain (HB101) harbouring the mobilization plasmid pRK2013 (Figarski and Helinski, PNAS 76, 648-1652, 1979). Transconjugant Agrobacterium is selected on LB plates with streptomycin and tetracyclin. Agrobacterium liquid cultures are prepared for the cocultivation experiments by growing them overnight at 28°C in a shaker in LB medium containing streptomycin and acetosyringone. The cultures are diluted 1:20 before using to give a density of approximately 1 x 10⁸ cells/ml.

Cotyledons from tomato cultivar UC82B, H722 or H902 seedlings are used in the co-cultivation procedures. Seeds from these cultivars are surface sterilized with 20% bleach and 0.1% Tween-20 for 15 minutes. The seeds are rinsed several times in sterile water and germinated on Vi MSO medium in magenta boxes. Then cotyledons from 2- week-old seedlings are removed, cut at both ends and precultured (upside down) on regeneration medium (MS 2.5 mg/1 BA, 2.5 mg/ 1 IAA) for one day. The precultured cotyledons are placed into the diluted overnight Agrobacterium culture for 15-20 min and placed onto regeneration media (MS, 5 mg/1 BA, 5 mg/1 IAA) for a co-cultivation period. After two days the cotyledons are placed on regeneration medium containing 500 mg/ml carbenicillin (to kill the Agrobacterium) and 10 mg/ml kanamycin for ten days. The cotyledons are transferred to regeneration medium with higher concentrations of kanamycin (first 50 μg/ml then 75 μg/ml) in subsequent transfers at 10 day intervals. Green shoots that regenerate from the cotyledons on 75 μg/ml kanamycin are transferred to rooting medium with 100 mg/ml kanamycin. Rooted shoots are transferred to soil and grown into plants in a growth room.

For analysis of transgenics, DNA is isolated from putative transgenics as described by Maniatis et al (eds., Molecular Cloning, a Laboratory Manual, Cold Spring Harbour Laboratory Press, New York, 1984). The presence of the anti-sense mannanase gene in the selected plants is determined by Southern blotting. Success with the ACC deaminase gene is a strong indication that this technique is valid and appropriate.

Claims

WE CLAIM:

1. A cDNA clone for endo-β-mannanase.

2. A cDNA clone according to claim 1 wherein the endo-β-mannanase is isolated from a fruit.

3. A cDNA clone according to claim 2 wherein the endo-β-mannanase is isolated from a fruit selected from the group including tomatoes, melons, peaches, oranges, cucumbers, and nectarines.

4. A cDNA clone according to claim 1 wherein the endo-β-mannanase is isolated from a tomato seed.

5. A cDNA clone which is at least 60% homologous to a cDNA clone for endo-β- mannanase.

6. An antisense sequence for endo-β-mannanase for inhibiting the expression of endo-β-mannanase.

7. An antisense sequence according to claim 6 wherein the endo-β-mannanase is obtained from a tomato seed.

8. An antisense sequence according to claim 6 wherein the endo-β-mannanase is obtained from a fruit.

9. An expression system comprising a cDNA clone for endo-β-mannanase and a promoter for expressing said cDNA clone for endo-β-mannanase.

10. An expression system comprising a cDNA clone which is at least 60% homologous to a cDNA clone for endo-β-mannanase and a promoter for expressing said cDNA clone.

11. A method for delaying the ripening of a fruit comprising the steps of: obtaining a cDNA clone for endo-β-mannanase produced by the fruit; and inhibiting the expression of endo-β-mannanase in the fruit.

12. A method according to claim 11 further including the steps of preparing an antisense sequence from the cDNA clone; and inserting the anti-sense sequence into the genome of the fruit.

13. A method according to claim 11 wherein the fruit is selected from the group including tomatoes, melons, peaches, oranges, cucumbers and nectarines.

14. A method for expediting the ripening of a fruit comprising the steps of: obtaining a cDNA clone for endo-β-mannanase and inserting a promoter for endo-β- mannanase into the genome of the fruit in order to over express the gene for endo- β-mannanase.

15. A method according to claim 14 wherein the fruit is selected from the group including tomatoes, melons, peaches, oranges, cucumbers and nectarines.

16. Use of a cDNA clone for endo-β-mannanase of a fruit for controlling the rate of ripening of the fruit.

17. Use according to claim 16 wherein the fruit is selected from the group comprising tomatoes, melons, peaches, oranges, cucumbers and nectarines.

18. Use according to claim 16 for delaying the ripening of a fruit.

19. Use according to claim 16 for expediting the ripening of the fruit.

20. Use of an anti-sense sequence for endo-β-mannanase for delaying the ripening of a fruit.