WO2008013450A1

WO2008013450A1 - A two-component system for seedless fruit development

Info

Publication number: WO2008013450A1
Application number: PCT/NL2007/050368
Authority: WO
Inventors: Arnaud Guillaume Bovy; Elias Gerardus Wilhelmus Maria Schijlen; Cornelis Hendricus De Vos; Gerrit Cornelis Angenent
Original assignee: Plant Research International B.V.
Priority date: 2006-07-24
Filing date: 2007-07-24
Publication date: 2008-01-31

Abstract

The application relates to the field of plant biotechnology. A two component system for making hybrid plants capable of producing seedless fruits is provided. Using the two component system, for example seedless tomatoes, peppers, aubergines and melons can 5 be made. Also provided is a method for inducible seed set.

Description

A two-component system for seedless fruit development

FIELD OF THE INVENTION

The present invention relates to the field of plant biotechnology and plant breeding. Provided are new methods for making seedless fruits, especially in plant species belonging to the family Solanaceae, such as tomato, pepper and aubergine and Cucurbitaceae, such as melon. These methods involve a modulation of the fiavonoid biosynthetic pathway, which is either blocked or downregulated by gene silencing of chalcone synthase gene(s). Alternatively the substrate pool for chalcone synthase is diverted by expression of an enzyme which utilizes the substrates, such as stilbene synthase. In addition a two-component system comprising plants capable of producing seedless fruits and the seedless fruits themselves are provided. In another aspect of the invention methods for inducible seed set and plants capable of inducible seed set are provided herein, as are fruits comprising such induced seeds.

BACKGROUND OF THE INVENTION

Flavonoids are plant secondary metabolites that are widespread throughout the plant kingdom. To date, more than 6000 flavonoids have been identified. Based on the structure of their basic skeleton, flavonoids can be divided into different classes, such as chalcones, fiavonols and anthocyanins (Figure 1). In nature, flavonoids are involved in many biological processes. For example they act as UV-light scavengers to protect against oxidative damage, as antimicrobial compounds to defend against pathogens and as pigments in fruits, flowers and seeds where they have a function in attracting pollinators and seed dispersers to facilitate reproduction (Koes et al., 1994, BioEssays 16, 123-132). Flavonoids are also present in pollen and pistils of many plant species (Wiermann, 1979, Regulation of Secondary Product and Plant Hormone Metabolism. Luckner and Schreiber, eds., Oxford: Pergamon Press, 231-239) and there is increasing evidence showing that flavonoids, at least in some plants, play a crucial role in fertility and sexual reproduction. For example, inhibition of fiavonoid production in Petunia plants, through antisense suppression of the gene encoding chalcone synthase (CHS), the first enzyme in the fiavonoid pathway (Figure 1), resulted not only in the inhibition of flower pigmentation, but also in male sterility (Van der Meer et al., 1992 Plant Cell 4, 253-62). Pollination experiments revealed that flavonoids in either the anther or pistil are essential for pollen tube growth, fertilization and subsequent seed set (Ylstra et al, 1992, Plant Physiol. 100, 902-907). After cross pollination, the sterile phenotype could be partly rescued by fiavonoids present in the wild type plant. In addition, in vitro experiments showed that flavonols, in particular kaempferol and quercetin are essential for pollen tube germination and growth in Petunia and maize (Ylstra et al., 1992 supra, 1996; Mo et al., 1992, Proc. Natl. Acad. ScL U.S.A. 89, 7213-7217). Further evidence for a role of fiavonoids in sexual reproduction is provided by the male sterile Petunia white anther (who) mutant, which could be complemented by the introduction of a functional CHS cDNA (Napoli et al., 1999 Plant Physiol. 120, 615-622).

In addition to (male) sterility, parthenocarpy, which is defined as the formation of seedless fruits in the absence of functional fertilisation (Gustafson 1942, Bot Rev 8, 599-654) is a desirable trait for several important crop plants. The production of seedless fruits can be of great value for consumers when directly eaten, but also for the processing industry. Besides this, parthenocarpy is advantageous when pollination or fertilisation is affected due to extreme temperatures. Unfortunately, mutations causing parthenocarpic fruits often have pleiotropic effects and can result in undesirable characteristics such as misshapen fruits (Varoquaux et al., 2000, Trends Biotech 18, 233-242).

To date, most seedless fruit production methods involve mutant plants comprising mutant alleles of particular genes (e.g. in tomato the pat- 2, pat- 3 and pat-4 mutants; see WO9921411) or transgenic approaches, wherein plant hormone production, transport or metabolism is altered (for example by expressing ispentenyl transferase or tryptophane oxygenase using promoters specific to the ovaries or developing fruit). Breeding methods using mutant alleles are often complicated due to complex multigenic systems and pleiotropic effects, as a result of which few seedless plants made by traditional breeding exist on the market to date. In addition, the transgenic approaches described to date have also suffered from complications, be it of a technical nature (for example requiring emasculation in order to produce seedless fruits) or of a public acceptance nature (e.g. involving 'terminator technology'). For a review on seedless fruit production methods and their limitations see Varoquaux et al. 2000 {supra). There still exists, therefore, a need for simple and efficient methods for producing seedless fruits. It is an object of the invention to provide such methods.

GENERAL DEFINITION "Seedless" refers herein to the complete absence of seeds in the (mature) fruit (i.e. "no seeds set") and/or a significant reduction in total seed number (i.e. "reduced seed set") and/or an arrest of seed development in the early stages, so that there is a significant reduction in the eventual number of fully developed seeds, whereby a significant reduction refers to a reduction to at least 40% of the wild type, preferably a reduction to at least 50%, 60%, 70%, 80%, 90%, 95% or 98%, most preferably a reduction to 100% of the wild type (i.e. completely seedless).

"Chalcone synthases" (also referred to as naringenin-chalcone synthase) refers to proteins of EC 2.3.1.74, capable of catalyzing the reaction: 3 malonyl-CoA + 4- coumaroyl-CoA = 4 CoA + naringenin chalcone + 3 CO₂ . "Resveratrol synthase" or "stilbene synthase" refers to proteins of EC 2.3.1.95, capable of catalyzing the reaction: 3 malonyl-CoA + 4-coumaroyl-CoA = 4 CoA + 3,4',5- trihydroxy-stilbene + 4 CO₂ .

"Flavonoid bio synthetic pathway" is shown in Figure 1. Chalcone synthase is the first enzyme in the pathway leading to chalcones, fiavonones, isofiavonoids, flavones and flavonols, and anthocyanins. Stilbene synthase diverts the substrates of chalcone synthase to form stilbenes.

"Hybrid plant(s)" refers to the Fl seeds, and Fl plant(s) grown from said seeds, produced from cross fertilization between two parental plants, i.e. obtained by pollination and fertilization between Pl (male parent) X P2(female parent) or between Pl (female parent) X P2 (male parent).

"Pollination dependent" refers to the development of seedless fruits only taking place when the plant's stigma is pollinated with live, compatible pollen. "Two component system" refers to hybrid seed/plant production system, wherein the hybrid plants are capable of producing seedless fruits, e.g. upon pollination without fertilization.

The term "nucleic acid sequence" (or nucleic acid molecule) refers to a DNA or RNA molecule in single or double stranded form, particularly a DNA encoding a protein or protein fragment according to the invention, or (in gene silencing approaches) a DNA or RNA molecule in sense and/or antisense orientation which is identical or substantially similar in sequence to an endogenous target gene (or gene transcript) or gene family of the host plant.

A "target gene" in gene silencing approaches is the gene or gene family (or one or more specific alleles of the gene) of which the endogenous gene expression is down- regulated or completely inhibited (silenced) when a chimeric silencing gene (or 'chimeric RNAi gene') is expressed and for example produces a silencing RNA transcript (e.g. a dsRNA or hairpinRNA capable of silencing the endogenous target gene expression). An "isolated nucleic acid sequence" refers to a nucleic acid sequence which is no longer in the natural environment from which it was isolated, e.g. the nucleic acid sequence in a bacterial host cell or in the plant nuclear or plastid genome. "Silencing" refers to a downregulation or completely inhibition of gene expression of the target gene or gene family. A "sense" RNA transcript is generally made by operably linking a promoter to a double stranded DNA molecule wherein the sense strand (coding strand) of the DNA molecule is in 5 ' to 3 ' orientation, such that upon transcription a sense RNA is transcribed, which has the identical nucleotide sequence to the sense DNA strand (except that T is replaced by U in the RNA). An "antisense" RNA transcript is generally made by operably linking a promoter to the complementary strand (antisense strand) of the sense DNA, such that upon transcription an antisense RNA is transcribed.

The terms "protein" or "polypeptide" are used interchangeably and refer to molecules consisting of a chain of amino acids, without reference to a specific mode of action, size, 3-dimensional structure or origin. A "fragment" or "portion" of a chalcone synthase or stilbene synthase protein may thus still be referred to as a "protein". An "isolated protein" is used to refer to a protein which is no longer in its natural environment, for example in vitro or in a recombinant bacterial or plant host cell. An enzyme is a protein comprising enzymatic activity.

The term "gene" means a DNA sequence comprising a region (transcribed region), which is transcribed into an RNA molecule (e.g. a primary transcript or a processed mRNA) in a cell, operably linked to suitable regulatory regions (e.g. a promoter). A gene may thus comprise several operably linked sequences, such as a promoter, a 5' leader sequence comprising e.g. sequences involved in translation initiation, a (protein) coding region (cDNA or genomic DNA) and a 3 'non-translated sequence comprising e.g. transcription termination sites.

A "chimeric gene" (or recombinant gene) refers to any gene, which is not normally found in nature in a species, in particular a gene in which one or more parts of the nucleic acid sequence are present that are not associated with each other in nature. For example the promoter is not associated in nature with part or all of the transcribed region or with another regulatory region. The term "chimeric gene" is understood to include expression constructs in which a promoter or transcription regulatory sequence is operably linked to one or more coding sequences or to an antisense (reverse complement of the sense strand) or inverted repeat sequence (sense and antisense, optionally separated by a nucleic acid spacer such as an intron, whereby the RNA transcript forms double stranded RNA or hairpin RNA upon transcription). "Expression of a gene" refers to the process wherein a DNA region, which is operably linked to appropriate regulatory regions, particularly a promoter, is transcribed into an RNA, which is biologically active, i.e. which is capable of being translated into a biologically active protein or peptide (or active peptide fragment) or which is active itself (e.g. in posttranscriptional gene silencing or RNAi). The coding sequence is preferably in sense-orientation and encodes a desired, biologically active protein or peptide, or an active peptide fragment. In gene silencing approaches, the DNA sequence is preferably present in the form of an antisense DNA or an inverted repeat DNA, comprising a short sequence of the target gene(s), or essentially similar thereto, in antisense, or in sense and antisense orientation. "Ectopic expression" refers to expression in a tissue in which the gene is normally not expressed. A "transcription regulatory sequence" is herein defined as a nucleic acid sequence that is capable of regulating the rate of transcription of a (coding) sequence operably linked to the transcription regulatory sequence. A transcription regulatory sequence as herein defined will thus comprise all of the sequence elements necessary for initiation of transcription (promoter elements), for maintaining and for regulating transcription, including e.g. attenuators or enhancers. Although mostly the upstream (5') transcription regulatory sequences of a coding sequence are referred to, regulatory sequences found downstream (3') of a coding sequence are also encompassed by this definition. As used herein, the term "promoter" refers to a nucleic acid fragment that functions to control the transcription of one or more genes, located upstream with respect to the direction of transcription of the transcription initiation site of the gene, and is structurally identified by the presence of a binding site for DNA-dependent RNA polymerase, transcription initiation sites and any other DNA sequences, including, but not limited to transcription factor binding sites, repressor and activator protein binding sites, and any other sequences of nucleotides known to one of skill in the art to act directly or indirectly to regulate the amount of transcription from the promoter. A "constitutive" promoter is a promoter that is active in most tissues under most physiological and developmental conditions. An "inducible" promoter is a promoter that is physiologically (e.g. by external application of certain compounds) or developmentally regulated.

A "tissue specific" or "tissue preferred" promoter mainly or only active in specific types of tissues or cells.

A "male organ/tissue specific promoter" is mainly or specifically active in one or more male reproductive tissues, such as the tapetum cells of the pollen grains, the microspores and/or the pollen grains.

A "female organ/tissue specific promoter" is mainly or specifically active in one or more female reproductive tissues, such as the stigma, the style, the ovary, the ovules and/or the pistil. "Early fruit development specific promoter" is a promoter mainly or specifically active during one or more stages of early fruit development.

As used herein, the term "operably linked" refers to a linkage of polynucleotide elements in a functional relationship. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a promoter, or rather a transcription regulatory sequence, is operably linked to a coding sequence if it affects the transcription of the coding sequence. Operably linked means that the DNA sequences being linked are typically contiguous and, where necessary to join two protein encoding regions, contiguous and in reading frame so as to produce a "chimeric protein". A "chimeric protein" or "hybrid protein" is a protein composed of various protein "domains" (or motifs) which is not found as such in nature but which a joined to form a functional protein, which displays the functionality of the joined domains (for example a DNA binding or a repressor domain leading to a dominant negative function). A chimeric protein may also be a fusion protein of two or more proteins occurring in nature. The term "domain" as used herein means any part(s) or domain(s) of the protein with a specific structure or function that can be transferred to another protein for providing a new hybrid protein with at least the functional characteristic of the domain.

The terms "target peptide" refers to amino acid sequences which target a protein to intracellular organelles such as vacuoles, plastids, preferably chloroplasts, mitochondria, leucoplasts or chromoplasts, the endoplasmic reticulum, or to the extracellular space (secretion signal peptide). A nucleic acid sequence encoding a target peptide may be fused (in frame) to the nucleic acid sequence encoding the amino terminal end (N-terminal end) of the protein or may replace part of the native amino terminal end of the protein.

A "nucleic acid construct" or "vector" is herein understood to mean a man-made nucleic acid molecule resulting from the use of recombinant DNA technology and which is used to deliver exogenous DNA into a host cell. The vector backbone may for example be a binary or superbinary vector (see e.g. US5591616, US2002138879 and WO9506722), a co-integrate vector or a T-DNA vector, as known in the art and as described elsewhere herein, into which a chimeric gene is integrated or, if a suitable transcription regulatory sequence is already present, only a desired nucleic acid sequence (e.g. a coding sequence, an antisense or an inverted repeat sequence) is integrated downstream of the transcription regulatory sequence. Vectors usually comprise further genetic elements to facilitate their use in molecular cloning, such as e.g. selectable markers, multiple cloning sites and the like (see below). A "host cell" or a "recombinant host cell" or "transformed cell" are terms referring to a new individual cell (or organism) arising as a result of at least one nucleic acid molecule, especially comprising a chimeric gene encoding a desired protein or a nucleic acid sequence which upon transcription yields an antisense RNA or an inverted repeat RNA (or hairpin RNA) for silencing of a target gene/gene family, having been introduced into said cell. The host cell may be any eukaryotic or prokaryotic cell e.g. a plant cell, microbial, insect or mammal (including human) cell. Preferably the host cell is a plant cell. The host cell may contain the nucleic acid construct as an extra- chromosomally (episomal) replicating molecule, or more preferably, comprises the chimeric gene integrated in the nuclear or plastid genome of the host cell. Included are any derivatives of the host cell, such as tissues, whole organism, cell cultures, explants, protoplasts, further generations (offspring from selfing or crossing), etc. derived from the cell which retain the introduced gene or nucleic acid.

The term "selectable marker" is a term familiar to one of ordinary skill in the art and is used herein to describe any genetic entity which, when expressed, can be used to select for a cell or cells containing the selectable marker. Selectable marker gene products confer for example antibiotic resistance, or herbicide resistance or another selectable trait such as a phenotypic trait (e.g. a change in pigmentation) or a nutritional requirements. The term "reporter" is mainly used to refer to visible markers, such as green fluorescent protein (GFP), eGFP, luciferase, GUS and the like. The term "ortholog" of a gene or protein refers herein to the homologous gene or protein found in another species, which has the same function as the gene or protein, but is (usually) diverged in sequence from the time point on when the species harbouring the genes diverged (i.e. the genes evolved from a common ancestor by speciation). Orthologs of the chalcone synthase or resveratrol synthase genes may thus be identified in other plant, animal, bacterial or fungal species based on both sequence comparisons (e.g. based on percentages sequence identity over the entire sequence or over specific domains) and/or functional analysis, such as enzyme activity assays. The terms "homologous" and "heterologous" refer to the relationship between a nucleic acid or amino acid sequence and its host cell or host organism, especially in the context of transgenic cells / organisms. A homologous sequence is thus naturally found in the host species (e.g. a tomato plant transformed with a tomato gene), while a heterologous sequence is not naturally found in the host cell (e.g. a tomato plant transformed with a sequence from potato plants). Depending on the context, the term "homolog" or "homologous" may alternatively refer to sequences which are descendent from a common ancestral sequence (e.g. they may be orthologs).

"Stringent hybridisation conditions" can be used to identify nucleotide sequences, which are substantially identical to a given nucleotide sequence. Stringent conditions are sequence dependent and will be different in different circumstances. Generally, stringent conditions are selected to be about 5°C lower than the thermal melting point (T_m) for the specific sequences at a defined ionic strength and pH. The T_m is the temperature (under defined ionic strength and pH) at which 50% of the target sequence hybridises to a perfectly matched probe. Typically stringent conditions will be chosen in which the salt concentration is about 0.02 molar at pH 7 and the temperature is at least 60⁰C. Lowering the salt concentration and/or increasing the temperature increases stringency. Stringent conditions for RNA-DNA hybridisations (Northern blots using a probe of e.g. lOOnt) are for example those which include at least one wash in 0.2X SSC at 63°C for 20min, or equivalent conditions. Stringent conditions for DNA-DNA hybridisation (Southern blots using a probe of e.g. lOOnt) are for example those which include at least one wash (usually 2) in 0.2X SSC at a temperature of at least 50⁰C, usually about 55°C, for 20 min, or equivalent conditions. See also Sambrook et al. (1989, infra) and Sambrook and Russell (2001, infra). "Sequence identity" and "sequence similarity" can be determined by alignment of two peptide or two nucleotide sequences using global or local alignment algorithms. Sequences may then be referred to as "substantially identical" or "essentially similar" when they (when optimally aligned by for example the programs GAP or BESTFIT using default parameters) share at least a certain minimal percentage of sequence identity (as defined below). GAP uses the Needleman and Wunsch global alignment algorithm to align two sequences over their entire length, maximizing the number of matches and minimises the number of gaps. Generally, the GAP default parameters are used, with a gap creation penalty = 50 (nucleotides) / 8 (proteins) and gap extension penalty = 3 (nucleotides) / 2 (proteins). For nucleotides the default scoring matrix used is nwsgapdna and for proteins the default scoring matrix is Blosum62 (Henikoff & Henikoff, 1992, PNAS 89, 915-919). Sequence alignments and scores for percentage sequence identity may be determined using computer programs, such as the GCG Wisconsin Package, Version 10.3, available from Accelrys Inc., 9685 Scranton Road, San Diego, CA 92121-3752 USA. Alternatively percent similarity or identity may be determined by searching against databases such as FASTA, BLAST, etc. For comparing sequence identity between sequences of dissimilar lengths, it is preferred that local alignment algorithms are used, such as the Smith Waterman algorithm (Smith TF, Waterman MS (1981) J. MoI. Biol 147(l);195-7), used e.g. in the EmbossWIN program "water". Default parameters are gap opening penalty 10.0 and gap extension penalty 0.5, using Blosum62 for proteins and DNAFULL matrices for nucleic acids. For comparing sequences with similar lengths, a global alignment algorithm is preferred.

In this document and in its claims, the verb "to comprise" and its conjugations is used in its non-limiting sense to mean that items following the word are included, but items not specifically mentioned are not excluded. In addition, reference to an element by the indefinite article "a" or "an" does not exclude the possibility that more than one of the element is present, unless the context clearly requires that there be one and only one of the elements. The indefinite article "a" or "an" thus usually means "at least one", e.g. "a cell" refers also to several cells in the form of cell cultures, tissues, whole organism, etc. Similarly, or "a fruit" or "a plant" also refers to a plurality of fruits and plants. It is further understood that, when referring to "sequences" herein, generally the actual physical molecules with a certain sequence of subunits (e.g. amino acids) are referred to. As used herein, the term "plant" includes the whole plant or any parts or derivatives thereof, such as plant organs (e.g. harvested or non-harvested storage organs, tubers, fruits, leaves, etc.), plant cells, plant protoplasts, plant cell tissue cultures from which whole plants can be regenerated, plant calli, plant cell clumps, and plant cells that are intact in plants, or parts of plants, such as embryos, pollen, ovules, fruits (e.g. harvested tissues or organs), flowers, leaves, seeds, tubers, clonally propagated plants, roots, stems, root tips and the like. Also any developmental stage is included, such as seedlings, cuttings prior or after rooting, etc.

As used herein, the term "variety" or "cultivar" means a plant grouping within a single botanical taxon of the lowest known rank, which can be defined by the expression of the characteristics resulting from a given genotype or combination of genotypes.

The term "allele(s)" means any of one or more alternative forms of a gene at a particular locus, all of which alleles relate to one trait or characteristic at a specific locus. In a diploid cell of an organism, alleles of a given gene are located at a specific location, or locus (loci plural) on a chromosome. One allele is present on each chromosome of the pair of homologous chromosomes. A diploid plant species may comprise a large number of different alleles at a particular locus. These may be identical alleles of the gene (homozygous) or two different alleles (heterozygous). The term "locus" (loci plural) means a specific place or places or a site on a chromosome where for example a gene or genetic marker is found.

DETAILED DESCRIPTION

It was surprisingly found that constitutive silencing of chalcone synthase expression

(using the CaMV 35S promoter) in tomato plants led to the development of tomato fruits having either no seeds or a significantly reduced number of seeds. As the seedless fruit phenotype could be complemented by crossing the transgenic plants with wild type plants in both directions (transgenic male X wild type female parent and vice versa, transgenic female X wild type male), the present inventors concluded that two components are essential for the development of seedless fruits: 1) the flavonoid biosynthetic pathway has to be blocked or down-regulated in both the male reproductive organs (the pollen grains or parts thereof) and the female reproductive organs (the pistil or parts thereof, such as the stigma, style and/or ovary or ovule tissue) in order to obtain seedless fruits and 2) pollination (using live, compatible pollen) is required for obtaining seedless fruits, i.e. pollination which does not result in fertilization of the ovules is necessary for triggering (seedless) fruit development.

Based on these findings, new methods for making seedless fruits have been devised, whereby different approaches can be used, all of which have the effect that the flavonoid biosynthetic pathway is blocked or down-regulated in (at least) both the male and female reproductive tissues of the parent plants used in the cross and that the cross (i.e. pollination using transgenic pollen from one parent, which is applied to the transgenic stigma of the other parent) results in Fl hybrid plants which develop fruits, but which do not initiate seed set within those fruits or which initiate a significantly reduced seed set (depending on the strength of the blockage/down-regulation of the flavonoid biosynthetic pathway in the reproductive tissues of the parent plants and the Fl hybrids).

Thus, when referring herein to "the flavonoid pathway being down-regulated or silenced" (in one or more tissues at certain times) this refers to the total flavonoids downstream of the enzyme chalcone synthase being significantly reduced (e.g. naringenine chalcone and quercetins) by any of the methods according to the invention.

This includes the actual down-regulation of the flavonoid pathway by gene silencing of the expression of the first enzyme in the pathway (chalcone synthase), but also the diversion of the pathway by overexpression of stilbene synthase.

As will be explained further below, these methods involve two separate transgenic parental plants, which are cross-fertilized to produce Fl hybrids. However, as shown in the Examples, also self-pollination, rather than cross-pollination, of a plant, in which the fiavonoid pathway is down-regulated or blocked in at least the male reproductive organs (e.g. the pollen grains or parts thereof such as the tapetum) and in at least the female reproductive organs (e.g. stigma/style), for example by using a promoter active in both male and female reproductive tissues, results in this plant developing fruits without seeds or with reduced numbers of seeds. However, for large scale production of plants capable of producing seedless fruits, a two component system is needed as the original transgenic plant can generally not be reproduced in large numbers. In addition, many plant species show hybrid vigor (increased growth and yield compared to the parental inbred lines), whereby it is anyhow desirable to develop hybrids.

Methods according to the invention

In one embodiment of the invention a two-component system for making seedless fruits, and hybrid seeds and hybrid plants capable of producing seedless fruits, is provided.

In its broadest terms, the method comprises the steps of:

(a) crossing a first parent plant (Pl) with a second parent plant (P2), wherein the first parent plant Pl comprises a first chimeric gene and the second parent plant P2 comprises a second chimeric gene integrated in the genome, and wherein both the first and second chimeric gene comprise a promoter active in plant cells operably linked to:

(i) a nucleic acid sequence which upon transcription is capable of silencing the endogenous chalcone synthase gene expression; or

(ii) a nucleic acid sequence encoding a (functional) stilbene synthase protein; and

(b) obtaining Fl hybrid seeds from said cross, and optionally further comprising the steps of

(c) growing the Fl hybrid plants from said seeds and harvesting the seedless fruits produced by said plants.

The two parent plants Pl and P2 may either both comprise a nucleic acid of (i) or of (ii), or alternatively one parent comprises a nucleic acid of (i) and the other parent comprises a nucleic acid of (ii), as it does not matter whether chalcone synthase expression is silenced or whether the flavonoid pathway is downregulated by diversion of the substrate used by chalcone synthase. The important feature is that the levels of flavonoids are downregulated in both parents, preferably at least in the male and female reproductive organs

For simplicity reasons, approaches (i) and (ii) are discussed separately below. The method also involves the generation of the transgenic parent plants used in the method. The parent plants are both fertile, i.e. they are not suffering from sterility and can be cross-fertilized with each another.

Recombinant parent plants Pl and P2 and gene silencing approaches (i) In one embodiment of the invention, two parent plants, Pl and P2, are provided each of which comprises a chimeric gene which, upon expression results in silencing of the expression of the endogenous chalcone synthase gene or gene family (in the cells or tissues in which the gene is expressed), and thereby in downregulation or silencing of the flavonoid biosynthetic pathway downstream of chalcone synthase (see Fig. 1).

Such plants can be made using known methods for constructing chimeric genes, vectors and plant transformation and regeneration.

Chalcone synthase genes (chs) generally exist as single copy or small gene families.

For example tomato plants contain CHSl and CHS2, see e.g. GenBank Accession numbers X55194 and X55195. Many chalcone synthase genes have been cloned to date, from a wide variety of plants and can be used according to the invention for making gene silencing constructs for plant transformation. Alternatively, homologs can be isolated (e.g. by PCR or using nucleic acid hybridization techniques and e.g. stringent hybridization conditions) or synthetic DNA sequences can be made de novo.

"Chalcone synthase genes" or "chalcone synthase (CHS) nucleic acid sequences" refer herein to nucleic acid sequences (cDNA, genomic DNA, RNA) which encode chalcone synthase proteins as defined below (CHS, see below). CHS nucleic acid sequences include variants which comprise at least 50, 60, 70, 75%, more preferably at least 80, 90, 95, 98, 99% or more nucleic acid sequence identity to SEQ ID NO: 1 and/or 3 and/or 5 (tomato), and/or to SEQ ID NO: 7 and/or 8 and/or 11 and/or 13 (pepper) and/or full length sequences comprising one or more of SEQ ID NO: 7, 8, 11 or 13, as determined using pairwise alignment using the GAP program. Such variants may also be referred to as being "essentially similar" to any one of SEQ ID NO: 1, 3 and/or 5 (tomato), and/or to SEQ ID NO: 7, 8, 11 and/or 13 (pepper). Also included are fragments of CHS nucleic acid sequences and variants. Fragments include parts of any of the above CHS nucleic acid sequences (or variants), which may for example be used as primers or probes or in gene silencing constructs. Parts may be contiguous stretches of at least 10, 15, 19, 20, 21, 22, 23, 25, 50, 100, 200, 300, 450, 500, 600, 700, 800, 900 or more nucleotides in length, of either the coding strand (sense strand) or the complementary strand (anti-sense strand,). Also included are therefore fragments of CHS nucleic acid sequences, whereby a fragment of at least about 20, 30, 40, 50 or 60 nucleotides in length comprises at least 50, 60, 70, 75%, more preferably at least 80, 90, 95, 98, 99% or more (100%) nucleic acid sequence identity to another fragment of a CHS nucleic acid sequence of about the same length.

It is clear that many methods can be used to identify, synthesise or isolate variants or fragments of CHS nucleic acid sequences, such as nucleic acid hybridization, PCR technology, in silico analysis and nucleic acid synthesis, and the like. Thus, an CHS- protein encoding nucleic acid sequence may be a sequence which is chemically synthesized or which is cloned from any plant species. In one embodiment of the invention a heterologous nucleic acid sequence is used to silence the endogenous chalcone synthase gene(s) of the host species to be transformed. For example, a petunia or pepper chalcone synthase gene (or variant or fragment thereof) may be used to silence chalcone synthase gene expression in transgenic tomato or aubergine plants. Alternatively, homologous chalcone synthase nucleic acid sequences may be used. For example a sequence originating from a particular plant species is reintroduced into said species. Thus, in another embodiment, the CHS DNA corresponds to, or is a modification/variant of, the endogenous CHS DNA of the species which is used as host species in transformation. Thus, a tomato CHS cDNA or genomic DNA (or a variant or fragment thereof) is preferably used to transform tomato plants. In addition (for regulatory approval and public acceptance reasons) the homologous or heterologous nucleic acid sequence may be operably linked to a transcription regulatory sequence, especially a promoter, which also originates from a plant species or even from the same plant which is to be transformed.

"Chalcone synthase" proteins are herein defined by their enzymatic activity (which can be tested in enzyme assays, see Examples) and/or by comprising an amino acid sequence identity to SEQ ID NO: 2 and/or SEQ ID NO: 4 and/or 6 (tomato), and/or to

SEQ ID NO: 9 and/or 10 and/or 12 and/or 14 (pepper) and/or full length proteins comprising one or more of SEQ ID NO: 9, 10, 12 or 14, of at least 50, 45, 50, 60, 70,

75%, more preferably at least 80, 85, 90, 95, 98, 99% or more, as determined using pairwise alignment using the GAP program, preferably using full length sequences.

Such variants may also be referred to as being "essentially similar" to any one of SEQ

ID NO: 2, 4 and/or 6, and SEQ ID NO: 9, 10, 12 and/or 14.

To generate parent plants comprising a chimeric gene, which upon expression results in silencing of the expression of an endogenous chalcone synthase gene or gene family, methods known in the art can be used.

"Gene silencing" refers to the down-regulation or complete inhibition of gene expression of one or more target genes, e.g. chalcone synthase genes, in a host cell or tissue. It is understood that in any transformation experiments a certain degree of variation in the phenotype of transformants is seen, normally due to position effects in the genome and/or due to copy number. Generally, "weak" and "strong" gene silencing plants are distinguished herein (all of which are embodiments of the invention), wherein "weak" gene silencing (RNAi) events refer to plants or plant parts wherein the endogenous target gene expression is reduced by about 15, 20 or 30% compared to the control tissue and "strong" gene silencing (RNAi) events refer to plants or plant parts wherein the endogenous target gene expression is reduced by at least about 70, 80, 90% or more compared to the control tissue (e.g. wild type). Silencing can be quantified by, for example, quantifying the transcript level of the target gene (e.g. using quantitative RT-PCR) and/or by determining and optionally quantifying the enzymatic activity of the target protein (in strong RNAi events the CHS activity is reduced to less than 10%, e.g. to less than or equal to about 2%, of the wild type activity) and/or by assessing and optionally quantifying the total fiavonoid levels or the levels of specific flavonoids in the target tissue (e.g. using HPLC analysis). For example in a strong gene silencing event total flavonoid levels may be reduced to less than 10%, preferably less than 5% or less than 2% of the wild type levels. See also the Examples. The degree of seedlessness of the fruits correlates with the strength of the RNAi events (strong RNAi events produce essentially 100% seedless fruits, while weak events produce significantly reduced numbers of seeds compared to wild type plants). The degree of seedlessness can therefore be fine tuned as desired.

The use of inhibitory RNA to reduce or abolish gene expression is well established in the art and is the subject of several reviews (e.g Baulcombe 1996, Stam et al. 1997,

Depicker and Van Montagu, 1997). There are a number of technologies available to achieve gene silencing in plants, such as chimeric genes which produce antisense RNA of all or part of the target gene (see e.g. EP 0140308 Bl, EP 0240208 Bl and EP

0223399 Bl), or which produce sense RNA (also referred to as "co-suppression"), see EP 0465572 Bl.

The most successful approach so far has however been the production of both sense and antisense RNA of the target gene ("inverted repeats"), which forms double stranded RNA (dsRNA) or a stem-loop structure (hairpin RNA, hpRNA) in the cell and silences the target gene(s) upon transcription from an upstream promoter. Methods and vectors for dsRNA and hpRNA production and gene silencing have been described in EP 1068311, EP 983370 Al, EP 1042462 Al, EP 1071762 Al and EP 1080208 Al.

A chimeric gene for plant transformation may, therefore, comprise a transcription regulatory region which is active in plant cells operably linked to a sense and/or antisense DNA fragment (or a complete nucleic acid sequence) of, or complementary or substantially similar to, a chs target gene or gene family.

Generally short (sense and anti-sense) stretches of the target gene sequence, such as 17, 18, 19, 20, 21, 22 or 23 nucleotides of coding and/or non-coding sequence of the target gene are sufficient. Longer sequences can also be used, such as at least 100, 200, 250, 500, 1000 or more nucleotides. Even DNA corresponding to, or being complementary to, the complete transcript RNA or mRNA may be used to make a sense and/or antisense construct. Preferably, the sense and antisense fragments/sequences are separated by a spacer sequence, such as an intron, which forms a loop (or hairpin) upon dsRNA formation.

In principle, any chs gene or gene family can be targeted. For example, one or several specific chs alleles may be silenced by choosing a nucleic acid region of their primary or mRNA transcripts specific for these alleles (see Byzova et al. Plant 2004 218: 379- 387 for allele specific silencing in an organ specific manner). Similarly, a whole gene family may be targeted for silencing by choosing one or more conserved regions of the transcripts for making the silencing construct. As mentioned above, the DNA region used in sense and/or antisense orientation does not need to be part of the coding region, but may also correspond to, or be complementary to, parts of the primary transcript (comprising a 5 ' and 3 ' untranslated sequence and introns) or to parts of the mRNA transcript (where any introns have been removed and a polyA tail has been added). It is understood that in a DNA sequence which corresponds to an RNA sequence the U is replaced by a T. It is also noted that in a chimeric gene which transcribes a dsRNA or hpRNA targeting capable of silencing chalcone synthase gene expression upon transcription in a host cell, the sense and antisense regions need not be of equal length and one region may be longer than the other (see also the Examples).

Thus, for example SEQ ID NO: 1, 3, 5, 7, 8, 11, 12 or 13, or variants thereof as described above, or fragments of any of these may be used to make a CHS gene silencing gene and vector and a transgenic plant in which one or more CHS genes are silenced in all or some tissues or organs. A convenient way of generating hairpin constructs is to use generic vectors such as pHANNIBAL and pHELLSGATE, vectors based on the Gateway® technology (see Wesley et al. 2004, Methods MoI Biol. 265:117-30; Wesley et al. 2003, Methods MoI Biol. 236:273-86 and Helliwell & Waterhouse 2003, Methods 30(4):289-95.), incorporated herein by reference. See also http://www.pi.csiro.au/rnai/ for other gene silencing vectors, such as inducible silencing vectors and vectors for silencing of multiple target genes.

By choosing conserved nucleic acid sequences all CHS gene family members in a host plant can be silenced, for example in tomato hosts both CHSl and CHS2 gene expression can be silenced or CHSl, CHS2 and CHS-B gene expression can be silenced. The silencing of all family members of a host plant is a preferred embodiment.

In one embodiment the promoter, which is operably linked to the sense and/or antisense nucleic acid sequence (to make a chimeric silencing / RNAi gene) is selected from a constitutive promoter, an inducible promoter, and (most preferably) a male organ specific promoter for the generation of one transgenic parent (Pl) and a female organ specific promoter for the generation of the other transgenic parent (P2). Optionally a 3' UTR may be operably linked to the 3' end of the chimeric gene, so that the operably linked DNA elements include promoter - CHS RNAi gene - 3 'UTR.

Preferred constitutive promoters include: the strong constitutive 35 S promoters or enhanced 35S promoters (the "35S promoters") of the cauliflower mosaic virus (CaMV) of isolates CM 1841 (Gardner et al, 1981, Nucleic Acids Research 9, 2871- 2887), CabbB-S (Franck et al, 1980, Cell 21, 285-294) and CabbB-JI (Hull and Howell, 1987, Virology 86,482-493); the 35S promoter described by Odell et al (1985, Nature 313, 810-812) or in US5164316, promoters from the ubiquitin family (e.g. the maize ubiquitin promoter of Christensen et al, 1992, Plant MoI. Biol. 18,675-689, EP 0 342 926, see also Cornejo et al. 1993, Plant Mol.Biol. 23, 567-581), the gos2 promoter (de Pater et al, 1992 Plant J. 2, 834-844), the emu promoter (Last et al, 1990, Theor. Appl. Genet. 81,581-588), Arabidopsis actin promoters such as the promoter described by An et al (1996, Plant J. 10, 107.), rice actin promoters such as the promoter described by Zhang et α/.(1991, The Plant Cell 3, 1155-1165) and the promoter described in US 5,641,876 or the rice actin 2 promoter as described in WO070067; promoters of the Cassava vein mosaic virus (WO 97/48819, Verdaguer et al 1998, Plant MoI. Biol. 37,1055-1067), the pPLEX series of promoters from Subterranean Clover Stunt Virus (WO 96/06932, particularly the S7 promoter), a alcohol dehydrogenase promoter, e.g., pAdhlS (GenBank accession numbers X04049, X00581), and the TRl' promoter and the TR2' promoter (the "TRl 'promoter" and "TR2'promoter", respectively) which drive the expression of the 1' and 2' genes, respectively, of the T-DNA (Velten et al, 1984, EMBO J 3, 2723-2730), the Figwort Mosaic Virus promoter described in US6051753 and in EP426641, histone gene promoters, such as the Ph4a748 promoter from Arabidopsis (PMB 8: 179-191), or others.

Alternatively, a promoter can be utilized which is not constitutive but rather is specific for one or more tissues or organs of the plant (tissue preferred / tissue specific, including developmentally regulated promoters). Most preferably, a male tissue specific promoter is used for generating one parent plant and a female tissue specific promoter is used for generating the other parent plant to be used in the methods.

Male tissue specific promoters include, for example, the tapetum specific promoters TA13 and TA29 from tobacco (US6562354; Koltunow et al. 1990, Plant Cell 2:1201- 1224; Seurinck et al. 1990 Nucleic Acids Res. 18: 3403), the tapetum specific promoter CA55 from Zea mays (EP570422), tapetum specific MS2 promoter from Arabidopsis (Aarts et al 1997, Plant J. 12:615-23), the tapetum specific A9 promoter from Arabidopsis (Paul et al. 1992, Plant MoI Biol 19: 611-622), the tapetum specific promoter BcA9 from Chinese cabbage (Lee et al. 2003, Plant Cell Rep. 22(4): 268-73), anther specific TAA promoters from wheat (Wang et al., 2002, Plant J. 30: 613-623), tapetum specific promoter from rice (e.g. PEl, T42, T72 from rice), a microspore development specific promoter such as NTM 19 from tobacco (EP790311) or a male germline specific promoter (e.g. LGCl from lily, WO9905281) or others may be used.

Female tissue specific promoters include stigma and/or style specific promoters such as the STIGl promoter from tobacco (Goldman et al. 1994, EMBO J 13: 2976-2984), the truncated SLG promoter from Brassica (Dzelzkalns et al. 1993, Plant Cell 5: 855-863) and the potato STS14 promoter (van Eldik et al. 1996, Plant MoI Biol 30: 171-176). Obviously, also the promoters from homologs or orthologs of these genes may be isolated and used. Also pistil specific promoters may be used or promoters active in both pistils and ovules. The DefH9 promoter (described in WO9828430) is active specifically in the placenta and in the ovules during early phases of flower development and may also be suitable. Additional promoters include, for example, promoters active in stigma and/or style and ovules. Alternatively, early fruit development specific promoters may be used. The skilled person can easily test various promoters for their specificity and suitability in the methods according to the invention. In addition, the specificity of promoters may be modified by deleting, adding or replacing parts of the promoter sequence. Such modified promoters can be operably linked to reporter genes in order to test their spatio-temporal activity in transgenic plants.

Another alternative is to use a promoter whose expression is inducible. Examples of inducible promoters are chemical inducible promoters, such as dexamethasone as described by Aoyama and Chua (1997, Plant Journal 11: 605-612) and in US6063985 or by tetracycline (TOPFREE or TOP 10 promoter, see Gatz, 1997, Annu Rev Plant Physiol Plant MoI Biol. 48: 89-108 and Love et al. 2000, Plant J. 21: 579-88). Other inducible promoters are for example inducible by a change in temperature, such as the heat shock promoter described in US 5,447, 858, by anaerobic conditions (e.g. the maize ADHlS promoter), by light (US6455760), by pathogens (e.g. EP759085 or EP309862) or by senescence (SAG12 and SAG13, see US5689042). Obviously, there are a range of other promoters available. Examples of other inducible promoters are the Adhl promoter which is inducible by hypoxia or cold stress, the Hsp70 promoter which is inducible by heat stress, and the PPDK promoter which is inducible by light. One preferred promoter is the ethanol-inducible promoter system, as described in Ait-ali et al. (2001, Plant Biotechnology Journal 1, 337-343), wherein ethanol treatment activates alcR, which in turn induces expression of the alc:35S promoter. See also Deveaux et al. (2003, The ethanol switch: a tool for tissue-specific gene induction during plant development. Plant J. 36, 918-930).

Optionally, the promoter-CHS^* RNAi gene may further comprise a 3'end transcription regulation signals ("3'end" or "3' UTR") (i.e. transcript formation and polyadenylation signals). Polyadenylation and transcript formation signals include those of, the nopaline synthase gene ("3' nos") (Depicker et al, 1982 J. Molec. Appl. Genetics 1, 561-573.), the octopine synthase gene ("3'ocs") (Gielen et al., 1984, EMBO J 3, 835-845) and the T-DNA gene 7 ("3' gene 7") (Velten and Schell, 1985, Nucleic Acids Research 13, 6981-6998), which act as 3 '-untranslated DNA sequences in transformed plant cells, and others.

The chimeric CHS silencing gene (i.e. the promoter operably linked to a nucleic acid sequence which upon transcription in a plant cell is capable of silencing the endogenous chalcone synthase gene expression) can be stably inserted in a conventional manner into the nuclear genome of a single plant cell, and the so- transformed plant cell can be used in a conventional manner to produce a transformed plant that has an altered phenotype due to CHS silencing in certain cells at a certain time. In this regard, a T-DNA vector, comprising a promoter operably linked to a sense and/or antisense CHS sequence (and optionally a 3'UTR), may be introduced into Agrobacterium tumefaciens and used to transform the plant cell, and thereafter, a transformed plant can be regenerated from the transformed plant cell using the procedures described, for example, in EP 0 116 718, EP 0 270 822, PCT publication WO84/02913 and published European Patent application EP 0 242 246 and in Gould et al. (1991, Plant Physiol. 95,426-434). The construction of a T-DNA vector for Agrobacterium mediated plant transformation is well known in the art. The T-DNA vector may be either a binary vector as described in EP 0 120 561 and EP 0 120 515 or a co-integrate vector which can integrate into the Agrobacterium Ti-plasmid by homologous recombination, as described in EP 0 116 718.

Preferred T-DNA vectors each contain a promoter operably linked to CHS silencing gene between T-DNA border sequences, or at least located to the left of the right border sequence. Border sequences are described in Gielen et al. (1984, EMBO J 3,835-845). Of course, other types of vectors can be used to transform the plant cell, using procedures such as direct gene transfer (as described, for example in EP 0 223 247), pollen mediated transformation (as described, for example in EP 0 270 356 and WO85/01856), protoplast transformation as, for example, described in US 4,684, 611, plant RNA virus- mediated transformation (as described, for example in EP 0 067 553 and US 4,407, 956), liposome-mediated transformation (as described, for example in US 4,536, 475), and other methods such as those described methods for transforming certain lines of corn (e. g., US 6,140, 553; Fromm et al., 1990, Bio/Technology 8, 833- 839; Gordon-Kamm et al., 1990, The Plant Cell 2, 603-618) and rice (Shimamoto et al., 1989, Nature 338, 274-276; Datta et al. 1990, Bio/Technology 8, 736-740) and the method for transforming monocots generally (PCT publication WO92/09696). For cotton transformation see also WO 00/71733, and for rice transformation see also the methods described in W092/09696, W094/00977 and W095/06722. For sorghum transformation see e.g. Jeoung JM et al. 2002, Hereditas 137: 20-8 or Zhao ZY et al. 2000, Plant MoI Biol.44:789-98). For tomato or tobacco transformation see also An G. et al, 1986, Plant Physiol. 81: 301-305; Horsch R.B. et al, 1988, In: Plant Molecular Biology Manual A5, Dordrecht, Netherlands, Kluwer Academic Publishers, pp 1-9; Koornneef M. et al., 1986, In: Nevins DJ. and R.A. Jones, eds. Tomato Biotechnology, New York, NY, USA, Alan R. Liss, Inc. pp 169-178). For potato transformation see e.g. Sherman and Bevan (1988, Plant Cell Rep. 7: 13-16).

Likewise, selection and regeneration of transformed plants from transformed cells is well known in the art. Obviously, for different species and even for different varieties or cultivars of a single species, protocols are specifically adapted for regenerating transformants at high frequency.

Besides transformation of the nuclear genome, also transformation of the plastid genome, preferably chloroplast genome, is included in the invention. One advantage of plastid genome transformation is that the risk of spread of the transgene(s) can be reduced. Plastid genome transformation can be carried out as known in the art, see e.g. Sidorov VA et al. 1999, Plant J.19: 209-216 or Lutz KA et al. 2004, Plant J. 37(6):906- 13.

Any plant may be a suitable host, such as monocotyledonous plants or dicotyledonous plants, but most preferably plants which produce fruits which would benefit from being seedless, such as tomato, pepper, cucumber, melon, aubergine (egg plant) and grape.

Preferred hosts are of the family Solanaceae, such as Lycopersicon ssp (reclassified as belonging to the genus Solarium), e.g. tomato (S. lycopersicum, e.g. cherry tomato, var. cerasiforme or current tomato, var. pimpinellifolium), tree tomato (S. betaceum, syn. Cyphomandra betaceae) and other Solanum species, such as aubergine/eggplant (Solanum melongena), pepino (S. muricatum), cocona (S. sessiliflorum) and naranjilla (S. quitoense). The family Solanaceae also includes peppers (Capsicum annuum, Capsicum frutescens), which are a preferred host. Other suitable hosts are pea (e.g. Pisum sativum; family Fabaceae) and various species bearing fleshy fruits (grapes, peaches, plums, strawberry, mango, papaya, etc.). Also Cucurbitaceae, such as melon (Citrullus lanatus, Cucumis melo) and cucumber (Cucumis sativus) and squashes and marrows (Cucurbita) are suitable hosts. Likewise Rosaceae are suitable hosts, such as apple, pear, plum, etc. Preferred host genera, therefore, are Solanum, Capsicum, Cucumis, Vitis, Citruttus or Cucurbita. Especially preferred host species are tomato species and pepper species, especially sweet pepper.

Thus, in one embodiment of the invention transgenic plants, Pl and P2, comprising a transcription regulatory element (especially a promoter as described above) operably linked to nucleic acid molecule a nucleic acid sequence which upon transcription is capable of silencing the endogenous chalcone synthase gene expression in the host cells.

Preferably, Pl and P2 comprises different promoters, most preferably the promoter in Pl is a male organ specific promoter and the promoter in P2 a female organ specific promoter. The promoter should be capable of expressing the RNAi gene in at least part of the pollen grains, and not the female parts, of the first parent and in at least part of the pistil, and not the male parts, in the second parent, so that the Fl hybrid plant produced from the Pl x P2 cross fertilization is capable of making pollen grains in which (or in part of which) CHS gene expression is silenced (and/or in which the flavonoid biosynthetic pathway is downregulated or blocked; see also embodiment ii, below) and pistils in which (or in part of which) CHS gene expression is silenced (and/or in which the flavonoid biosynthetic pathway is downregulated or blocked; see also embodiment ii, below).

In a preferred embodiment Pl and P2 are homozygous for the transgene, so that the Fl hybrid seeds each comprise one chimeric CHS-RNAi gene from each parent.

Homozygous transgenic plants can be made by selfing the plants. It is noted that Pl and

P2 themselves, as well as any parts thereof, such as the pollen grains or other tissues and organs, as well as derivatives of Pl and P2, are also an embodiment of the invention. If Pl and P2 are hemizygous, then preferably the Fl hybrid seeds/plants are selected which comprise a transgene copy from each parent.

The method for making Fl hybrid seeds capable of producing seedless fruits comprises crossing Pl with P2. As most preferably Pl comprises a male organ specific CHS^* silencing gene, while P2 preferably comprises a female organ specific CHS silencing gene, it is preferred to contact the pollen of P2 with the stigma of Pl. The pollen is then capable of fertilizing the eggs of Pl and produces Fl hybrid seeds. The cross- pollination of Pl with P2 can be carried out manually or by planting Pl and P2 in proximity of each other, e.g. in parallel rows or by interplanting (planting mixtures of Pl and P2 seeds or seedlings).

Pl and P2 plants may be selected to be "strong" transgenic events, as the Fl hybrids will then also show a strong silencing phenotype (capable of developing 99% or 100% seedless fruits).

The Fl hybrid seeds are then collected (e.g. from the Pl plants) by known methods. A plurality of Fl hybrid seeds, each comprising one CHS-RNAi gene from Pl and from P2 are also an embodiment of the invention. These can be bagged and sold for the production of seedless fruits. Preferably, the Fl hybrid seeds and plants obtained therefrom are "strong", i.e. preferably capable of producing 99% or 100% seedless fruits.

To produce seedless fruits, the Fl seeds are sown and grown. Upon flowering and pollination these plants are capable of triggering fruit development without fertilization, but in a pollination dependent manner. The seedless fruits are produced in step (c) generally by self-pollination of the Fl stigma with the Fl pollen of the same plant and/or same flower. However, it is also possible to use a different transgenic pollen

(from another transgenic plant), as long as the fiavonoid pathway is downregulated in (at least part of) the pollen grains. For example, the Fl stigma can also be pollinated with pollen of one of the parent plants (e.g. Pl), or with pollen of another Fl plant.

The seedless fruits are harvested and can be sold fresh, processed or stored, as desired. A plurality of harvested seedless fruits is a further embodiment herein. These fruits can be distinguished from other plants by the presence of the transgenes, although these will preferably not be expressed in the fruit tissue (i.e. the promoters are preferably not active in developed fruit tissue, such as fruit flesh or peel). However, if the promoters are active, it is shown in the Examples that the CHS silencing will likely not affect taste of the fruits in any negative way, even though some phenotypic changes may be seen. For certain purposes, it may even be desired that the seedless fruits are more solid and/or smaller, as was seen in tomato when chalcone synthase was constitutively silenced (see Examples). Such solid seedless fruit are also encompassed herein.

Recombinant parent plants Pl and P2 and stilbene synthase expression (ii)

In another embodiment, the substrate of CHS is depleted or reduced by diversion and the fiavonoid biosynthetic pathway downstream of CHS is downregulated as a consequence.

This method comprises transgenic plants wherein the (functional) enzyme stilbene synthase (STS) is expressed. Thus, in one embodiment of the invention any nucleic acid sequence encoding a stilbene synthase protein, or protein variant, or functional protein fragment, may be used for making a chimeric gene, vector and transformed plant or plant cell, using an expression vector as described further below.

Any STS nucleic acid sequence (cDNA, genomic DNA, RNA) encoding a STS protein or functional protein fragment may be used (referred herein to as "STS nucleic acid sequence"). A "stilbene synthase (STS) protein" refers to functional protein have stilbene synthase enzymatic activity (as can be tested in enzyme assays as known in the art) (see e.g. Rolfs et al, Plant Cell Reports 1, 83-85, 1981) and structurally by the percentage sequence identity over the entire length. Stilbene synthase proteins have a sequence identity of at least 40%, or more, over their entire length to SEQ ID NO: 16 and/or 18 (grape and peanut proteins, respectively), such as but not limited to at least 43%, 45%, 50%, 55%, 56%, 58%, 60%, 70%, 75%, 80%, 85%, 90%, 95%, 98%, 99%, 99.5,%, 99.8% or more at the amino acid sequence level, as determined using pairwise alignment using the GAP program (with a gap creation penalty of 8 and an extension penalty of 2). Such variants, and functional fragments of these, may also be referred to as being "essentially similar" to SEQ ID NO: 16 or 18. Preferably proteins having some, preferably 1-10, 20 or more amino acids added, replaced or deleted without significantly changing the protein activity are included in this definition. For example conservative amino acid substitutions within the categories basic (e. g. Arg, His, Lys), acidic (e. g. Asp,Glu), nonpolar (e. g. Ala, VaI, Trp, Leu, He, Pro, Met, Phe, Trp) or polar (e. g. GIy, Ser, Thr, Tyr, Cys, Asn, GIn) fall within the scope of the invention as long as the enzymatic activity of the stilbene synthase protein is not significantly, preferably not, changed or at least not reduced, e.g. when compared with the activity of SEQ ID NO: 16 or 18. In addition non-conservative amino acid substitutions fall within the scope of the invention as long as the activity of the stilbene synthase protein is not changed significantly, preferably not changed or at least not reduced, e.g. when compared with the activity of SEQ ID NO: 16 or 18. Also functional stilbene synthase protein fragments and active chimeric stilbene synthase proteins are encompassed herein. Protein fragments may be fragments of at least about 5, 10, 20, 40, 50, 60, 70, 90, 100, 150, 160, 200, 220, 230, 250, 300, or more contiguous amino acids. Also, the smallest protein fragment which retains activity in vivo in plants (and is therefore functional) is also provided. A nucleic acid sequence encoding such a fragment may be use to generate a transgenic plant as described.

Also included are STS nucleic acid sequences (e.g. SEQ ID NO: 15 and 17) and variants and fragments thereof, such as nucleic acid sequences hybridizing to STS nucleic acid sequences (e.g. to SEQ ID NO: 15 or 17) under stringent hybridization conditions as defined. Variants of STS nucleic acid sequences include nucleic acid sequences which have a nucleic acid sequence identity to SEQ ID NO: 15 and/or 17 (grape and peanut STS) of at least 50% or more, preferably at least 55%, 60%, 70%, 80%, 90%, 95%, 99%, 99.5%, 99.8% or more, as determined using pairwise alignment using the GAP program using full lengths sequences. Such variants may also be referred to as being "essentially similar" to any one of SEQ ID NO: 15 or 17. Also fragments are provided. Fragments may be contiguous stretches of at least 10, 15, 19, 20, 21, 22, 23, 25, 50, 100, 200, 450, 500, 700, 800, 900 or more nucleotides in length and preferably encode a functional STS protein or protein fragment. Preferably the STS nucleic acid sequences are of plant origin (i.e. they naturally occur in plant species) or are modified plant sequences.

It is clear that many methods can be used to identify, synthesise or isolate nucleic acid sequences encoding STS proteins or variants or fragments thereof, such as nucleic acid hybridization, PCR technology, in silico analysis and nucleic acid synthesis, and the like. Thus, an STS-protein encoding nucleic acid sequence may be a sequence which is chemically synthesized or which is cloned from any organism, preferably plant sequences are used. Orthologs of the grape and peanut STS genes (SEQ ID NO: 15 and 17) may for example be isolated from other plants or identified in silico.

Furthermore, for optimized in-planta expression the codon usage of the STS-encoding nucleic acid sequence is, in one embodiment, adapted, most preferably to the preferred codon usage of the host species which is to be transformed. In a preferred embodiment any of the above STS DNA sequences (or variants) are codon-optimized by adapting the codon usage to that most preferred in the host genus or preferably the host species (Bennetzen & Hall, 1982, J. Biol. Chem. 257, 3026-3031; Itakura et al., 1977 Science 198, 1056-1063.) using available codon usage tables (e. g. more adapted towards expression in tomato, aubergine, melon, etc.). Codon usage tables for various plant species are published for example by Ikemura (1993, In "Plant Molecular Biology Labfax", Croy, ed., Bios Scientific Publishers Ltd.) and Nakamura et al. (2000, Nucl. Acids Res. 28, 292.) and in the major DNA sequence databases (e.g. EMBL at Heidelberg, Germany). Accordingly, synthetic DNA sequences can be constructed so that the same or substantially the same proteins are produced. Several techniques for modifying the codon usage to that preferred by the host cells can be found in patent and scientific literature. The exact method of codon usage modification is not critical for this invention. Other modification, which may optimize expression in plants and/or which make cloning procedures easier may be carried out, such as removal of cryptic splice sites, avoiding long AT or GC rich stretches, etc. Such methods are known in the art and standard molecular biology techniques can be used.

The embodiments described above for CHS silencing generally also apply to STS (over)expression in host species and to the production of producing Fl hybrids capable of producing seedless fruits. Thus, as above, a promoter (constitutive, inducible, or preferably male or female tissue specific, or a promoter active in early fruit development) is operably linked to a nucleic acid sequence encoding a functional STS protein and a host plant is transformed and regenerated using known methods. As mentioned, the functionality of an STS protein (or variant or fragment) can be tested by using enzymatic assays known in the art. Similarly, in vivo activity in the transgenic plant tissue can be tested by analysing the amount of stilbene produced upon expression of the nucleic acid sequence encoding the protein. This may for example be done using HPLC analysis.

The method involves crossing a first parent Pl, with a second parent P2, wherein both parent plants produce a functional STS upon in the tissue where the promoter is active. The substrate for the endogenous CHS protein is thereby diverted into the production of stilbene and the CHS activity in vivo is significantly reduced, in correlation with the increased STS production and activity. Thereby, the fiavonoid biosynthetic pathway downstream of CHS is downregulated or blocked.

Preferably two "strong" events are selected as Pl and P2, although in this embodiment "strong" refers to the ability to produce large amounts of active protein upon expression. As above, the method involves producing an Fl hybrid plant which is capable of producing seedless fruits.

As mentioned above, methods (i) and (ii) may also be combined. For example, a parent Pl or P2 of method (i) may be crossed with a parent Pl or P2 of method (ii) in order to produce Fl hybrid plants capable of developing seedless fruit. It is, therefore, not required that both parents comprise either a CHS-RNAi gene or a nucleic acid sequence encoding a STS protein, but one parent may comprise an CHS-RNAi gene while the other parent comprises a nucleic acid sequence encoding a STS protein.

Inducible seed set

In a different embodiment of the invention a method for inducing the development of seeds (seed set) is provided. Seed set can be induced in the plants described above by inducing the fiavonoid pathway in at least one of those plants. Upon induction of the fiavonoid pathway at the right time the plant (which normally produces seedless fruits) will develop 'normal' fruits, i.e. comprising seeds. "Induced seed set" preferably refers to the induced plant producing at least 50%, preferably at least 60%, 70%, 80%, 90% or most preferably 100% of the number of seeds which a wild type plant would produce.

The method comprises the steps of: (a) Providing a transgenic plant wherein the flavonoid pathway is downregulated at least in the male and/or female reproductive organs or parts thereof (at one or more timepoints), i.e. providing a plant which is capable of producing seedless fruits; and (b) introducing one or two further chimeric genes into said plant, wherein the chimeric genes comprise an inducible promoter active in plant cells, operably linked to a nucleic acid sequence encoding a protein which stimulates the flavonoid pathway; and optionally

(c) inducing said promoter by contacting the plant (one or more times) with a suitable amount of inducer for a suitable period of time.

The plants of step (a) may, for example, be any of the plants described above, i.e. Pl or P2 or Fl hybrids, all as described above.

The further chimeric genes of (b) may be introduced by further transformation of a plant already comprising a chimeric CHS-RNAi or STS gene, or by crossing transgenic plants of (a) with a transgenic plant transformed with a chimeric gene of (b).

Alternatively, a plant which comprises several chimeric genes can be made de novo. For example several chimeric genes can be present on a single transformation vector or be co -transformed at the same time using separate vectors and selecting transformants comprising all of the chimeric genes.

Thus, eventually a transgenic plant is provided herein which comprises the following genetic elements: a promoter active in plant cells operably linked to a chimeric CHS-RNAi gene or to a chimeric stilbene synthase gene, whereby the flavonoid biosynthetic pathway is silenced or diverted, such that the plant produces seedless fruits in a pollination dependent manner (e.g. as described above); and (a) an inducible promoter operably linked to a first nucleic acid sequence encoding a protein which stimulates (upregulates) the flavonoid biosynthetic pathway, and optionally a further inducible promoter operably linked to a second nucleic acid sequence encoding a protein which stimulates the flavonoid biosynthetic pathway, or, alternatively

(b) an inducible promoter operably linked to a nucleic acid sequence encoding a functional CHS protein, with the provision that, if the flavonoid biosynthetic pathway is downregulated by a chimeric CHS-RNAi gene, the nucleic acid sequence encoding the functional CHS protein is not silenced by the CHS-RNAi gene (i.e. is not sensitive to CHS-RNAi silencing). The inducible production of the CHS protein will then complement the CHS silencing and result in a stimulation of the flavonoid biosynthetic pathway.

The protein which stimulates (upregulates) the flavonoid biosynthetic pathway is preferably a MYC type R family transcription factor and/or a MYB type Cl family transcription factor (Dooner et al. 1991, Ann Rev Genet 25: 173-199; Bovy et al. 2002, Plant Cell 14: 2509-2526). For example, the maize LC transcription factor (the protein of SEQ ID NO: 20, encoded by SEQ ID NO: 19) and/or a MYB type Cl transcription factor (the protein of SEQ ID NO: 22, encoded by SEQ ID NO: 21) may be expressed under control of an inducible promoter. Alternatively homo logs or orthologs of these may be isolated or identified and expressed. An "LC transcription factor" and a "Cl transcription factor" refers herein to proteins comprising at least 40%, 50%, 60, 70, 80%, more preferably at least 85%, 90%, 95%, 98%, 99% or more amino acid sequence identity to SEQ ID NO: 20 and 22, respectively, as well as functional fragments of these. In addition hybrid proteins are encompassed herein, for example comprising one or more domains from an LC transcription factor and from a C 1 transcription factor, or comprising two or more domains from different LC transcription factors, or from different Cl transcription factors. For example, the bHLH domain of different LC proteins may be combined or the bHLH domain of an LC protein may be combined with one or more domains of a Cl protein. Such a hybrid protein would be advantageous if it is in itself sufficient for upregulating the flavonoid pathway (see below).

The transcription factors are "functional" if they are capable of upregulating the flavonoid biosynthetic in vivo, upon overexpression of the nucleic acid sequences encoding the proteins, variants or fragments. See Bovy et al. (2002, supra) showing that in LC/C1 overexpressing in tomato a >100 fold induction of fiavonoid bio synthetic genes takes place (e.g. of CHS, flavanone-3 -hydroxylase, dihydrofolate reductase and a 5 to 15 fold induction of fiavone synthase, anthocyanine synthase, flavonol-3- glucosyltransferase and fiavonol-3-glucoside-rhamnosyl transferase) relative to the wild type. The capability to upregulate the fiavonoid pathway can, therefore, be determined by generating a transgenic plant or plant tissue and determining e.g. by HPLC, if fiavonoid production is increased or by qRT-PCR if gene expression is induced (for example chalcone synthase expression).

Preferably both a nucleic acid sequence encoding an LC transcription factor and a Cl transcription factor are expressed in the same plant. The transcription of the transcription factors is preferably regulated from separate inducible promoters (which may be identical or different), but it is also envisaged to express the genes from a single promoter, e.g. as a polycistronic transcript.

Codon usage of the nucleic acid sequences encoding the LC and Cl transcription factors may be adapted to the host genus or species which is to be transformed (as described above for STS encoding sequences).

Suitable inducible promoters are described herein above.

The inducing stimulus is preferably contacted with the plant prior to and/or during flowering, for example prior to and/or during and/or shortly after pollination. The amount and type of inducing stimulus of course depends on the promoter used, but the optimal amount and timing, as well as the optimal method of contact and formulation, can be determined using routine experimentation.

In an alternative embodiment a CHS nucleic acid sequence (encoding a functional CHS protein) is over-expressed in an inducible manner, so that the CHS downregulation is complemented after induction and the fiavonoid biosynthetic pathway is stimulated. The CHS nucleic acid sequence used in the downregulation of the endogenous CHS gene should be sufficiently different from the CHS nucleic acid sequence used in over- expression, in order to avoid silencing of the CHS transgene (encoding the functional CHS protein). Thus, preferably a sufficiently different CHS nucleic acid sequence is used for complementation, for example a heterologous sequence from a different plant species than the host CHS sequence.

One can easily identify and make a CHS-RNAi which does not silence another (heterologous) CHS allele, but which does silence the CHS family of the transgenic host. Whether the CHS-RNAi gene does or does not silence the chimeric CHS^* gene encoding the functional CHS protein can be tested using routine methods, especially by plant transformation.

Obviously, also other genetic elements may be introduced into any of the plants, such a as other genes conferring phenotypic input or output traits (e.g. herbicide resistance genes, disease resistance genes, etc.), antibiotic resistance genes, genetic elements for marker gene removal (cre/lox, flp/frt, etc.) and others.

It is understood that whole plants, seeds, fruits, cells, tissues and progeny of any of the transformed plants described above are encompassed herein and can be identified by the presence of the transgene or transgenes in the DNA, for example by PCR analysis using total genomic DNA as template and using specific PCR primer pairs. Also "event specific" PCR diagnostic methods can be developed, where the PCR primers are based on the plant DNA flanking the inserted chimeric gene, see US6563026. Similarly, event specific AFLP fingerprints or RFLP fingerprints may be developed which identify the transgenic plant or any plant, seed, tissue or cells derived there from.

Also fresh and processed food or feed compositions comprising parts (e.g. fruits) of the transgenic plants are encompassed herein. This includes fruit and/or vegetable juices, extracts, homogenates, etc. obtained using the plants or plant parts.

It is understood that the transgenic plants according to the invention preferably do not show non-desired phenotypes, such as yield reduction, enhanced susceptibility to diseases or undesired architectural changes (dwarfing, deformations) etc. and that, if such phenotypes are seen in the primary transformants, these can be removed by normal breeding and selection methods (crossing / backcrossing / selfing, etc.). Any of the transgenic plants and parts thereof described herein may be homozygous or hemizygous for the transgene(s).

SEQUENCES SEQ ID NO 1: tomato chalcone synthase 1 (CHSi) mRNA SEQ ID NO 2: tomato chalcone synthase 1 (CHSl) protein SEQ ID NO 3: tomato chalcone synthase 2 (CHS2) mRNA SEQ ID NO 4: tomato chalcone synthase 2 (CHS2) protein SEQ ID NO 5: tomato chalcone synthase B (CHS-B) mRNA SEQ ID NO 6: tomato chalcone synthase B (CHS-B) deduced amino acid sequence SEQ ID NO 7: pepper chalcone synthase 1 (CHSl) partial mRNA 5' SEQ ID NO 8: pepper chalcone synthase 1 (CHSl) partial mRNA 3'

SEQ ID NO 9: pepper chalcone synthase 1 (CHSl) deduced amino acid sequence; partial N-terminal

SEQ ID NO 10: pepper chalcone synthase 1 (CHSl) deduced amino acid sequence; partial C-terminal

SEQ ID NO 11: pepper chalcone synthase 2 (CHS2) partial mRNA 5'

SEQ ID NO 12: pepper chalcone synthase 2 (CHS2) deduced amino acid sequence; partial N-terminal

SEQ ID NO 13: pepper chalcone synthase B (CHS-B) partial mRNA 5'

SEQ ID NO 14: pepper chalcone synthase B (CHS-B) deduced amino acid sequence; partial N-terminal

SEQ ID NO 15: grape resveratrol synthase 1 mRNA SEQ ID NO 16: grape resveratrol synthase 1 protein

SEQ ID NO 17: peanut resveratrol synthase mRNA

SEQ ID NO 18: peanut resveratrol synthase protein

SEQ ID NO 19: Zea mais MYC type LC transcription factor mRNA

SEQ ID NO 20: Zea mais MYC type LC transcription factor protein SEQ ID NO 21 : Zea mais MYB type C 1 transcription factor mRNA

SEQ ID NO 22: Zea mais MYB type Cl transcription factor protein

SEQ ID NO 23: pHEAP vector sequence used for CHS RNAi FIGURE LEGENDS

Figure 1. Schematic overview of the flavonoid biosynthesis pathway in plants. The pathway normally active in tomato fruit peel, leading to flavonol production, is indicated by solid arrows. Abbreviations: CHS, chalcone synthase; STS, stilbene synthase; CHI, chalcone isomerase; F3H, fiavanone hydroxylase; FNS, fiavone synthase; IFS, isoflavone synthase; FLS, flavonol synthase, F3'H, flavonoid-3' -hydroxylase; F3'5'H, flavono ids', 5 '-hydroxylase; DFR, dihydrofiavonol-4-reductase; ANS, anthocyanidin synthase.

Figure 2. Schematic drawing of the tomato Chs RNAi construct.

Transgene expression was under control of the CaMV double 35S promoter (Pd35S ) and terminated by the Agrobacterium tumefaciens nos terminator (Tnos). An inverted repeat was generated by cloning a sense Chs-1 cDNA fragment (801 bp) followed by the full length cDNA sequence encoding tomato Chs-1 in anti-sense orientation.

Figure 3. Comparison of flavonoid levels between Wt and Chs RNAi tomato. a. Total flavonoid levels in leaf extracts of different CHS RNAi transgenic lines. b. HPLC chromatograms obtained from non hydro lyzed fruit peel extracts of wild type (upper panel) and Chs RNAi plants (lower panel). In the control plant the major compounds found are naringenin-chalcone (NC) and the favonol rutin (R) . c. Percentage of flavonoids in the fruit peel of Chs RNAi plants (lines 34, 39, 44 and 24) relative to the fruit peel of control tomatoes. Mean control values (mg/kg FW): naringenin chalcone = 212.5, sd 66.5; quercetin derivatives (rutin + rutin apioside) = 80.7, sd 11.0.

Figure 4. Quantitative Real Time PCR analysis.

Steady state mRNA levels of tomato Chs\ and Chs2 relative to the housekeeping gene L33 (encoding tomato ribosomal protein L33) were measured in fruit peel extracts of RNAi (34, 39, 44 and 24) and control lines. Expression levels in control were set to 100%.

Figure 5. CHS enzyme activity. a). Autoradiography scan of extraction of CHS assays developed on cellulose plates in CAW. Left panel showing CHS activity of tomato fruit peel in different Chs RNAi and control lines. Right panel showing CHS activity in fruits obtained after (reciprocal) crossings between WT and Chs RNAi. b). Densitometric scans of profiles of selected assays. The peak representing the CHS reaction product Naringenin (NAR) peak is indicated.

Figure 6. Tomato fruit weight in grams (black bars) and fruit size at equatorial cross section in mm (grey bars). Values represent are mean values ± sem. Control (WT), n=10; line 34, n=4; line 44, n=15; line 24 and 39, each n=14.

Figure 7. Overview of different pheno types found in Chs RNAi tomatoes compared with wild type. a) Typical ripe wild type tomato fruits are shiny and orange red, in contrast to dull, smaller and more reddish Chs RNAi fruits (b, c and d; line 24, 34 and 39 respectively). Fruits derived from flowers that were pollinated with wild type pollen (arrow) grew to normal size and obtained their shininess (d). Transgenic s lines 44 and 34 yielded extreme small fruits and 'fruit caves' when compared to wild type (e). In parthenocarpic Chs RNAi fruits seed development was disturbed (g) or totally absent (h and i), whereas wild type fruits had a normal seed set (f)

Figure 8. Electron microscopy photograph of epidermal cells of red ripe tomato fruits. A and C: surface view; B and D: cross section.

Wild type (A and B) fruits contain conical shaped cells on the epidermal surface, whereas in Chs RNAi (C and D) fruits the epidermal cell layer is disturbed (absence of conical shapes and 'empty cells').

Figure 9: Histochemical staining of WT and Chs RNAi (line 24) pollen tube growth in carpels 2 days after pollination. Fertilized carpels were stained with aniline blue to specifically stain callose present in growing pollen tubes. A, E, and I are WT carpels crossed with WT pollen; B, F, and J are WT carpels x Chs RNAi pollen; C, G, and K represent Chs RNAi x WT; finally D, H, and L are Chs RNAi self crossings. A-D show pollen at the stigma. Note in D, callose in the pollen tubes is still visible at the stigma (arrow), indicating inhibited growth. E-H, proliferation of pollen tube growth in the middle of the style, except for H which is only ¹A of the way down the style from the stigma. No pollen tubes were visible in the middle of the styles from Chs RNAi selfed plants. I-L, pollen tube growth at the base of the style, except K which grew only 9/10 of the way down the style. In K, the tips of the pollen tubes are swollen (arrow). Pollen tubes are not visible at the base of the style in K (not shown) or L. All micrographs are the same magnification.

The following non-limiting Examples describe the use of CHS genes for modifying plant phenotypes. Unless stated otherwise in the Examples, all recombinant DNA techniques are carried out according to standard protocols as described in Sambrook et al. (1989) Molecular Cloning: A Laboratory Manual, Second Edition, Cold Spring Harbor Laboratory Press, and Sambrook and Russell (2001) Molecular Cloning: A Laboratory Manual, Third Edition, Cold Spring Harbor Laboratory Press, NY; and in Volumes 1 and 2 of Ausubel et al. (1994) Current Protocols in Molecular Biology, Current Protocols, USA. Standard materials and methods for plant molecular work are described in Plant Molecular Biology Labfax (1993) by R.D.D. Croy, jointly published by BIOS Scientific Publications Ltd (UK) and Blackwell Scientific Publications, UK.

EXAMPLES

Example 1 - Experimental protocol

1.1 Plasmid construction

A full length cDNA encoding tomato (Lycopersicon esculentum) naringenin-chalcone synthase- 1 (Chsl; X55194) was obtained from a cDNA library of tomato fruits. Two oligonucleotides CHS-3 'BamHl (GGATCCACTAAGCAGCAACAC) and CHS-5\Sα/I (GTCTCGTCGACATGGTCACCGTGGAGGA) were used to introduce a BamHl restriction site at the 3 'end and a Sail site at the 5 'end of the Chs-1 sequence. The PCR product was digested and ligated as a BamHl / Sail fragment into pFLAP50, a pUC derived vector containing a fusion of the double CaMV 35S promoter (Pd35S) and the Agrobacterium tumefaciens nos terminator (Tnos). The resulting plasmid was designated as pHEAP- 02.

To create an inverted repeat construct a sense cDNA fragment was cloned between the promoter sequence and the anti-sense Chs-1. Therefore an 801 bp fragment was obtained by PCR amplification using two primers with restriction sites for BgIII (forward primer CCCAGATCTATGGTCACCGTGGAGGAGTA; reverse primer CCCAGATCTTCACGTAAGGTGTCCGTCAA) The BgHI digested PCR fragment was cloned in the BamHI digested plasmid HEAP-02 resulting in the plasmid pHEAP-17. The Pd35S-Chsl inverted repeat- Tnos construct was transferred as a Pacl/Ascl fragment into pBBC90, a derivative of the plasmid pGPTV-KAN ^¹¹^ and the final binary plasmid was designated pHEAP-20 (SEQ ID NO: 17).

1.2 Plant transformation The plasmid pHEAP-20 was transferred to A. tumefaciens strain COR308 by the freeze-thaw method (Gynheung, et al, 1988, Binary vectors. In: Plant Molecular Biology Manual, S. B. Gelvin, R.A. Schilperoort, and D.P.S. Verma, eds (Dordrecht, The Netherlands: Kluwer Academic Publishers, pp. A3/1-A3/19). The Agrobacterium mediated tomato (hypocotyls) transformation (cv. Money maker) was performed according to the standard protocol (Fillati et al., 1987, Bio.Technol. 5, 726-730). Kanamycin-resistant shoots were transferred to the greenhouse to grow on rock wool. The transgenic status of the plants was confirmed by PCR specific for the introduced gene and by Southern blot hybridisation (DIG labeling, Roche). Plants were allowed to self-pollinate in order to give fruits and offspring.

For further analysis (HPLC, DNA and RNA) fruits were harvested when visually ripe. From each plant at least three fruits were pooled for extraction to minimize sample variation. The fruit peel (approximately 2mm consisting of cuticula, epidermis and sub- epidermis) was separated from the flesh tissue (i.e. columella; jelly parenchyma and seeds excluded) and immediately frozen in liquid nitrogen. Beside fruit material also young leaves were collected and frozen in liquid nitrogen to store at -80°C for later use.

1.3 HPLC analysis

Flavonoid content was determined both as glycosides and aglycons by preparing non- hydro lyzed and acidic-hydro lyzed extracts respectively. Non hydrolyzed extracts were prepared in 75% aqueous methanol using 15 minutes of sonication. Subsequent HPLC of the extracted flavonoids was performed with a gradient of 5 to 50% acetonitrile in 0.1% formic acid. Absorbance spectra and retention times of eluting peaks were compared with those of commercially available flavonoid standards (Apin chemicals, Abingdon, UK). Analysis of flavonoids in the extracts was performed by reverse phase HPLC (Phenomenex Luna 3μm C 18, 150 * 4,50 mm column, at 40°C) with photodiode array detection (Waters 996).

1.4 RNA isolation

Total RNA was isolated from tomato fruits as described previously (Bo vy et al, 1995). After DNAse-I treatment and RNeasy column purification (Qiagen GmbH), the total RNA yield was measured by Absorption at 260 nm. To determine the RNA quality, a small amount (1 μg) of each sample was evaluated on a 1% TAE agarose gel.

1.5 RT-PCR gene expression analysis

Real time quantitative (RT) PCR analysis was performed to test the effect of the anti- sense transcript on the endogenous gene expression levels of CHS. TaqMan sequence detection primers were designed based on the published Chs sequences from tomato (Chsl; X55194 and Chs2; X55195) by use of the SDS 1.9 Software (Applied Biosystems). The designed primer combinations (Table I) were synthesised by Applied Biosystems. Two μg total RNA was used for cDNA synthesis using Superscript II reverse transcriptase (Invitrogen) in a 100 μl final volume according to the standard protocol. The ABI 7700 sequence detection was used to measure the gene expression of each gene in triple in the presence of the fluorescent dye SYBR-Green. The expression of the genes Chsl and Chs2 was related to the constitutively expressed gene encoding ribosomal protein L33 (TC85035). Calculations of the expression in each sample were carried out according to the standard curve method (PE Applied Biosystems).

Table 1. Oligonucleotides used for TaqMan analysis.

1.6 Enzyme assay

Naringenin (NAR) was from Roth (Karlsruhe, Germany). [2-¹⁴C]malonyl-CoA (spec, act. 53 mCi/mmol) was from Hartmann Analytic (Braunschweig, Germany). A- coumaryol-CoA was a gift from W. Heller (Neuherberg, Germany). [4a, 6, 8-¹⁴C] NAR) was prepared as described in Martens et al. (2006) using recombinant chalcone synthase and chalcone isomerase. Radioactivity incorporated in labeled substrate was quantified by scanning sample aliquots after migration on cellulose plates (Merck, Darmstadt, Germany) using a bio-Imaging Analyzer Fuji BAS FLA 2000 (Raytest, Straubenhardt, Germany) and by direct scintillation counting (LKB Wallac 1214 Rackbeta, PerkinElmer Wallac, Turku, Finland).

Proteins were extracted from grounded fruit tissue as follows: 200 mg tissue was homogenized with 100 mg sea sand, 200 mg Dowex 200-400 mesh (aquil. 0,1 M Tris- HCl, pH 7,5) in 1 ml 0,1 M Tris-HCl, pH 7,5 containing 20 mM sodium ascorbate. After two centrifugation steps at 10.000 x g (Sorvall RMC 14, Du Pont Nemours GmbH, Bad. Nauheim, Germany) for 5 min at 4°C the resulting supernatant was directly used for CHS assays. Protein concentration was determined according to Bradford (1976), using BSA as a standard. For each sample two independent preparations were performed. Standard assays for CHS was performed in a final volume of 200 μl and contained : 140 μl 0.1 M Tris-HCl, pH 7.5, 50 μl crude extract (8 - 22 μg protein), 5 μl [2-¹⁴C] malonyl-CoA (1,5 nmol; -1800 Bq) and 5 μl 4-coumaroyl-CoA (1 nmol). After incubation reactions were stopped and extracted twice with 100 μl ethyl acetate. The pooled EtOAc phase from each assay was directly subjected to scintillation counter for quantification or chromatographed on cellulose plates with either CAW (chloroform : acetic acid : water; 50 / 45 / 5) or 15% acetic acid. For each enzyme preparation CHS assays were performed in triplicate. Labelled products were localized and quantified by scanning the plates as above described. Product identification was done by co- chromatography with authentic samples.

1.7 in vivo pollen tube growth

Mature closed flowers were emasculated and pollinated. Two days after pollination (dap), pistils were harvested and incubated overnight at 60°C in 1 M KOH. After rinsing with water, pistils were transferred to a microscope slide and stained with 0.005% aniline in 50% glycine. A coverslip was placed on top and pressed gently. Callose in the pollen tubes was visualized by UV light on a Zeiss Axioskop microscope, photographed using 400 ASA film. Slides were scanned with an AGFA duoscan scanner.

1.8 Cryo-Scanning Electron Microscopy

Small samples of material were dissected from fresh fruits, mounted on a stub, and subsequently frozen in liquid nitrogen. The samples were further prepared in an Oxford Alto 2500 cryo-system (Catan, Oxford, UK), and then analysed in a JEOL JSM-6330F field emission electron scanning microscope (JEOL, Tokyo, Japan). The frozen samples were fractionated inside the cryo-system for cross-views.

Example 2 - Results

2.1 RNAi strategy down-regulates Chs gene expression in tomato In order to down-regulate the fiavonoid biosynthesis in tomato, we introduced a Chs-1 RNAi gene construct (Figure 2) using Agrobacterium-meάmtQά plant transformation. This Chs RNAi construct was expressed under control of the constitutive enhanced cauliflower mosaic virus (CaMV) 35S promoter, and therefore it was expected that the transgene effect would not be restricted to the tomato fruit only, but would also influence the fiavonoid pathway in other parts of the tomato plant.

2.2 Biochemical analysis of fiavonoid levels

In total 15 PCR-positive transgenic Chs RNAi TO plants were used for a first biochemical analysis. Based on HPLC analyses of leaf as well as fruit peel extracts, transgenic plants showing various degrees of reduced total fiavonoid levels could be identified (Figure 3A).

From these primary transformants, four single-copy transgenic lines with strongly decreased fiavonoid levels were selected for further analysis. Of each plant, three cuttings were propagated and from each individual cutting a sample was collected encompassing at least three ripe fruits. These samples from transgenic and wild type tomato plants were analyzed for fiavonoid content using HPLC. The main flavonoids accumulating in WT tomato fruits are naringenin-chalcone and the flavonol rutin (quercetin-3-rutinoside). In fruits, these flavonoids are predominantly produced in the peel, since the flavonoid pathway is inactive in flesh tissue (Muir et al., 2001 Nature Biotech. 19, 470-474; Bovy et al., 2002, Plant Cell. 14, 2509-2526). In peel extracts of Chsϊ fruits, a large decrease was observed in the levels of both rutin and naringenin- chalcone when compared with extracts from control plants (Figure 3B+C). According to the percentage of naringenin chalcone and quercetin derivatives relative to control plants, lines 24, 39 and 44 were regarded as lines with a "strong" phenotype (< 5% of total flavonoids left), and line 34 as a "weak" phenotype (approximately 30 % left).

2.3 Chs gene expression analysis In tomato, Chs is a member of a small multi-gene family comprising at least two genes with high sequence similarity (Yoder et al., 1994, Euphytica 79, 163-167). Gene- specific oligonucleotides were designed to discriminate between both Chsl and Chs 2 mRNA. The samples used for biochemical analysis were also used to measure the expression of the endogenous Chsl and Chs2 genes by real time quantitative RT PCR. The constitutively expressed tomato gene encoding ribosomal protein L33 was used as internal standard. Expression of this gene was found to be constitutive in DNA micro- array experiments with Chs RNAi and wild type fruit peel, as well as during different stages of fruit ripening (data not shown). Compared to wild type, the "strong" Chs RNAi lines showed a large decrease in expression levels of both Chsl and Chs 2 (Figure 4). In contrast, a relatively small decrease in Chs expression was found in fruits of line 34, the "weak" phenotype. For all lines, the observed decreases in gene expression levels correlated well with the biochemical data.

A similar decrease was found in CHS enzyme activity (Figure 5). The strong Chs RNAi lines appeared to have the lowest CHS activity and product levels (reduced to 2 % of wild type values) whereas most remaining activity was found in line 34. In fruits derived from reciprocal crossing the presence of CHS activity was related to the (WT) maternal genotype, giving rise to fruit peel tissue.

2.4 Phenotvpic characterization of Chs RNAi tomatoes The Chs RNAi tomato plants were phenotypically similar to wild-type with respect to the vegetative tissues. However, all the "strong" Chs RNAi plants showed a delayed fruit development and yielded smaller fruits (Figure 6). In addition, ripe fruits derived from Chs RNAi plants were reddish and the colour of their peel was dull (Figure 7b and c), in contrast to wild ripe fruits that are more orange-red and shiny (Figure 7a). The more intense red colour of Chs RNAi fruits was most probably due to the reduction in the levels of the yellow-pigmented naringenin-chalcone, normally present at high levels in epidermal cells of the ripening fruit (Hunt and Baker, 1980, Phytochem. 19, 1415-1419). Interestingly, chalcone isomerase (Chi) overexpressing fruits display reduced naringenin chalcone levels, amore intense red colour and a dull appearance as well (Muir et al, 2001, supra). This suggests a relationship between flavonoids and fruit dullness. To investigate the dull appearance of the fruit peel in more detail, red wild type and Chs RNAi fruits were subjected to electron microscopy analysis. The epidermal cell layer of the wild type fruits consisted of intact cells with a typical conical shape (Figure 8a and b) which normally confers the properties of higher light absorption and velvet sheen. The absence of this conical cell surface in Chs RNAi fruits could be an explanation for the dullness, as was described for the Antirrhinum majus (mixta) and the Petunia (mybPhl) mutants (Noda et al., 1994, Nature 369, 661-664; van Houwelingen et al., 1998, Plant J. 13, 39-50; MoI et al., 1998, Trends in Plant Sci. 3, 212-217) in which the fainter petal colours also resulted from flattening of the epidermal cells. Furthermore, the fruit surface of the Chs RNAi plants consisted of dead epidermal cells that were empty, as shown by the scanning EM freeze-fraction cross view (Figure 8c and d). A similar collapsed 'flat tyre' appearance of epidermal cells was observed in petals of the Petunia shrivelled up (shp) mutants and led to a drastic change of flower color (MoI et al., 1998, supra).

2.5 Parthenocarpic fruit development

A more detailed investigation of the fruits revealed that the majority of the transgenic lines tested produced parthenocarpic fruits, containing no seed at all ("strong" pheno types) or arrested seed set at early stages of development ("weak" pheno types). Within each line, the fruit phenotype was quite constant, but between different transgenic lines the phenotype varied considerably. Like in many parthenocarpic plants (Falavigna et al., 1978, Genet Agraria 32, 159-160; Abad et al., 1989, Sci Hort 38, 167- 192) undersized, misshapen or hollow fruits were also found in Chs RNAi tomato. At the end of the growing season the Chs RNAi phenotype appeared to be more extreme in terms of very small fruit size for some lines when compared to wild type fruits (Figure 7d). Fruits of some Chs RNAi plants contained no jelly and were completely filled with "flesh". An overview of all the phenotypes found is shown in Figure 7.

2.6 Plant pollen tube growth and fertility Since fiavonoids were shown to play an essential role in pollen germination and pollentube growth in Petunia and maize (Ylstra et al., 1992, 1996; Mo et al., 1992, supra) we investigated whether these processes were affected in Chs RNAi plants as well. Fertilized carpels of wild type, Chs RNAi (line 24), and reciprocal crossed plants were histochemically stained specific for callose present in growing pollen tubes 2 days after pollination. In wild type self pollinated plants pollen tubes reached the ovules 2 days after pollination (Figure 9b, f and j). Within the same time period, pollen tubes of Chs RNAi self pollinated plants germinated, but did not grow well. In these plants, callose staining was clearly visible in the stigma and absent further down the style, indicating an inhibited pollen tube growth (Figure 9 d, h and 1). Pollen tube growth could be (partially) rescued by crossing wild type and Chs RNAi plants in both directions. Pollen tube growth in wild type female plants pollinated with Chs RNAi pollen reached the base of the style 2 days after pollination (Figure 9b, f and j). The reciprocal crossing, i.e. Chs RNAi pistils pollinated with wild type pollen, was less efficient (figure 9c, g and k). Here, the pollen tubes grew only to about 9/10 of the way down the style and some of the tube tips were swollen.

Several cross-fertilised flowers were allowed to give fruits to see if the rescued pollen tube growth was able to yield fruits with normal seed production (Table 2).

Table 2 - Fruit seed set (mean ± s.d.) subsequent to crossings between wilde type (WT) and CHS RNAi tomato line 39 and 44.

(? x c?) WT x WT 39 ) c WT 44 x WT WT x 39 WT x 44

Seeds per fruit 121 ± 47 35 ± 8 57 ± 22 121 ± 41 129 ± 22

Fruits (n) 4 6 7 8 5

Seed production was fully rescued when wild type female plants were pollinated with

Chs RNAi pollen. Wild type pollen was also able to give rise to seed production in Chs RNAi female plants, however this was less efficient. Interestingly, the size of fruit obtained after Chs RNAi flowers were pollinated with wild type pollen increased to normal and the fruits gained their velvet sheen, although they were still more reddish, due to the absence of naringenin chalcone (Figure Id). It is likely that seed set and fruit shininess result from complex interactions between more development factors. Apparently flavonoids present in wild type pollen are sufficient to give rise to seed set, and possibly they also trigger directly or indirectly signals involved in fruit peel formation. The wild type flowers that were pollinated with Chs RNAi pollen gave rise to fruits that were indistinguishable from normal wild type fruits.

2.7 Transgene stability and offspring

A few offspring plants (Fl) obtained from transgenic line 34 and several obtained from crossings of Chs RNAi with wild type plants were selected for further evaluation of inheritance stability of the transgene. The low flavonoid phenotype was shown to segregate with the Chs RNAi transgene in all plants tested (n=8). The segregation of the transgene could already be seen in light stressed seedlings. Non transgenic seedlings accumulated anthocyanins in stems and leaf axis when grown under high light conditions and became purple. In contrast, in Chs RNAi transgenic seedlings the inhibition of flavonoid biosynthesis resulted in the absence of anthocyanins. Also in fruits of the transgenic offspring the total flavonoid levels remained very low (less than 1% of non transgenic fruits, data not shown).

2. 8 Discussion

The above experiments show that the flavonoid pathway in tomato can be efficiently down-regulated by RNAi-mediated suppression of chalcone synthase gene expression. In fruits, this led to a strong decrease in expression of both Chsl and Chs2 genes and CHS activity. As a consequence, an up to 99% reduction in total flavonoids was measured. This was mainly due to reduced levels of naringenin chalcone and rutin, the predominant flavonoids in tomato peel. Chs RNAi fruits showed a dull appearance and altered fruit colour, due to aberrations in the epidermal cell layers. Also pollen development was hampered, resulting in a strongly reduced seed set. Surprisingly, all strong Chs RNAi lines yielded parthenocarpic fruits, which can be used to devise novel methods for making seedless fruit. A relation between flavonoids and parthenocarpic fruit development has never been described, although it is well-known that flavonoids present in the sculptured cavities of the pollen exine, the so called pollen coat (Edlund et al, 2004, Plant Cell. 16, S84- S97), are essential for pollen development, germination and pollen tube growth and hence play an essential role in plant reproduction. For example, mutation of the two Chs genes in maize as well as the Petunia white anther (who) mutant resulted in white, flavono id- lacking pollen that showed to be sterile. Furthermore, Petunia plants harbouring a complete block of fiavonoid production due to anti-sense Chs or sense Chs co-suppression had white flowers and were male sterile (Van der Meer et al., 1992; Napoli et al, 1999).

Especially flavonoids belonging to the class of flavono Is have been shown to have strong stimulatory effects on pollen development, germination, pollen tube growth and seed set (Ylstra et al., 1992, 1996; Mo et al., 1992). The inability of pollen from the sterile wha mutant to germinate normally could be complemented by the introduction of a functional Chs transgene (Napoli et al., 1999) or by flavonol addition. When applied to Wt stigmas, tube growth of Chs deficient sterile pollen and seed set could be partly rescued (Mo et al., 1992; Taylor et al., 1992). This has led to the assumption that Chs deficient pollen lacks factors that are required for pollen tube growth and that Wt stigmas can functionally complement with these factors (Mo et al., 1992).

Herein it was observed that pollen tube growth was strongly inhibited in self-pollinated Chs RNAi tomato plants, leading to parthenocarpic fruit development. Since both male and female Chs RNAi parents were hemizygous, and hence produced gametes segregating for the transgene, the observed effects on pollen tube growth, seed set and parthenocarpic development are determined by parental tissues. The tapetum, and consequently pollen wall assembly, as well as the maternal tissues such as stigma and the style can play crucial roles in functional pollen rehydration, polarization and pollen tube migration into the stigma. Control of these processes likely requires constant interaction between pollen tube and stigma.

Pollen tube growth and seed set was fully rescued when CHS deficient pollen was applied on WT stigmas. The reciprocal crossing (WT pollen on Chsi stigmas) resulted in only a partial rescue of pollen tube growth and seed set, indicating that in Chs RNAi tomatoes, fertilisation is mainly diminished due to the lack of fiavonoids in the female reproductive organ.

Pollination appeared to be required for parthenocarpic fruit development in Chs RNAi lines, since in the absence of pollination no fruits were obtained. This suggests that pollination is required and sufficient to trigger fruit setting and that fertilization and subsequent seed set are key determinants for normal fruit development and expansion.

Several tomato mutant genotypes resulting in parthenocarpic fruit growth have been described of which pat (Mazzucato et al., 2003, Sex plant reprod. 16, 157-164), pat2 (Philouze et al., 1978, Tomato Genet Coop Rep. 28, 12-13) and pat3, pat4 (Nuez et al., 1986, Z Pflanzenzuchtung 96, 200-206) are the best characterised. In contrast to the Chs RNAi tomato described herein, parthenocarpic fruit development of all these pat mutants is independent of pollination. At least some of these mutants (pat2, pat3 and pat4) contain increased GA levels in their parthenocarpic fruits, suggesting that the ability of these mutants to develop parthenocarpic fruits is due to alterations in GA metabolism (Fos et al., 2001, Physiologia plantarum. I l l, 545-550).

Herein a new method to obtain pathenocarpic (seedless) fruit is described. For a successful commercial application of this technology it is an essential prerequisite that these parthenocarpic fruits have a good taste. To address whether or not the parthenocarpic phenotype dramatically affects fruit taste, the levels of the most important taste and flavour-related tomato metabolites (sugars, organic acids and 16 volatiles (Yilmaz, 2001, Turk J Agric For 25, 149-155; Ruiz et al., 2005, J Sci Food Agric. 85, 54-60) in Chsi and Wt tomatoes were measured. For both Wt and Chsi fruits, the levels of flavour-related volatile compounds fell well within the variation observed in a collection of 94 commercially available tomato cultivars (Tikunov et al, 2005, Plant Physiol., 139, 1125-1137). Similar results were obtained for sugars (sucrose, fructose and glucose) and organic acids (citric acid and malic acid) (results not shown), suggesting that these parthenocarpic tomatoes potentially have a normal tomato-like taste.

Claims

1. A method for making seedless fruits, and hybrid seeds and plants capable of producing seedless fruits, comprising the steps of: (a) cross-fertilizing a first parent plant with a second parent plant, wherein the first parent plant comprises a first chimeric gene and the second parent plant comprises a second chimeric gene integrated in the genome, and wherein both said first and second chimeric gene comprise a promoter active in plant cells operably linked to: i) a nucleic acid sequence which upon transcription silences endogenous chalcone synthase gene expression; or ii) a nucleic acid sequence encoding a functional stilbene synthase protein; and

(b) obtaining Fl hybrid seeds from said cross-fertilization, and wherein said first chimeric gene comprises a male reproductive-tissue specific promoter and wherein said second chimeric gene comprises a female reproductive- tissue specific promoter.

2. The method according to claim 1, further comprising the steps of: (c) growing the Fl hybrid plants from said seeds and harvesting the seedless fruits produced by said plants.

3. The method according to claim 1 or 2, wherein said chimeric gene is homozygous in said first and second parent plant and wherein said Fl hybrid seeds and Fl hybrid plants and fruits comprise said first chimeric gene from the first parent and said second chimeric gene from the second parent.

4. The method according to any one of the preceding claims, wherein the first and the second parent plant are both fertile plants.

5. The method according any one of the preceding claims, wherein the promoter of said first chimeric gene is a tapetum specific promoter and the promoter said second chimeric gene is a stigma and/or style or pistil specific promoter.

6. The method according to any one of the preceding claims, wherein said nucleic acid sequence in step i) comprises a sense and/or antisense sequence of a nucleic acid sequence encoding a protein comprising at least 60% amino acid sequence identity to SEQ ID NO: 2, 4 or 6; or wherein said stilbene synthase protein in step ii) comprising at least 60% amino acid sequence identity to SEQ ID NO: 16 or 18.

7. The method according to any one of the preceding claims, wherein the plants, seeds and fruits belong to the genus Solarium, Capsicum, Cucumis, Vitis,

Citruttus or Cucurbita, and preferably are tomato, pepper, aubergine or melon plants, seeds and fruits.

8. Fl hybrid seeds or Fl hybrid plants or seedless fruits obtainable by the method of any one of the preceding claims.

9. A method according to any one of the preceding claims wherein said first or second parent plant and/or said Fl hybrid plant further comprises a chimeric gene comprising an inducible promoter active in plant cells, operably linked to a nucleic acid sequence encoding a protein which stimulates the flavonoid pathway, preferably an Lc and/or Cl transcription factor or a chalcone synthase protein.

10. A method for inducing seed set in a plant which produces seedless fruits due to silencing of endogenous chalcone synthase gene expression or due to overexpression of stilbene synthase in one or more reproductive tissues, said method comprising the steps of a) introducing a chimeric gene into said plant genome, wherein said chimeric gene comprises an inducible promoter active in plant cells operably linked to a nucleic acid sequence encoding a protein which stimulates the flavonoid pathway, preferably an Lc and/or C 1 transcription factor or a chalcone synthase protein; and b) contacting said plant with the inducer of said promoter at one or more time points prior to or during flowering.

11. A transgenic plant comprising integrated in its genome a first and second chimeric gene, wherein said first chimeric gene comprises a promoter active in plant cells operably linked to: i) a nucleic acid sequence which upon transcription silences endogenous chalcone synthase gene expression; or ii) a nucleic acid sequence encoding a stilbene synthase protein; and wherein said second chimeric comprises an inducible promoter active in plant cells operably linked to a nucleic acid sequence encoding a MYC type R family transcription factor and/or a MYB type Cl family transcription factor which stimulates the fiavonoid pathway.

12. A kit of parts comprising a first transgenic parent plant or seed and a second transgenic parent plant or seed, wherein the first parent plant or seed comprises a first chimeric gene and the second parent plant or seed comprises a second chimeric gene integrated in the genome, and wherein both said first and second chimeric gene comprise a promoter active in plant cells operably linked to: i) a nucleic acid sequence which upon transcription silences endogenous chalcone synthase gene expression; or ii) a nucleic acid sequence encoding a stilbene synthase protein; and wherein said first chimeric gene comprises a male reproductive tissue specific promoter and wherein said second chimeric gene comprises a female reproductive tissue specific promoter.

13. The kit according to claim 12, further comprising Fl hybrid plants or seeds obtained from cross-fertilizing said first and second transgenic plant.

14. A plant or seed comprising a first chimeric gene and a second chimeric gene integrated in the genome, and wherein both said first and second chimeric gene comprise a promoter active in plant cells operably linked to: i) a nucleic acid sequence which upon transcription silences endogenous chalcone synthase gene expression; or ii) a nucleic acid sequence encoding a stilbene synthase protein; and wherein said first chimeric gene comprises a male reproductive tissue specific promoter and wherein said second chimeric gene comprises a female reproductive tissue specific promoter.

15. Seedless fruit obtained from the plant according to claim 14.