CA2146113A1 - Genetic moderation of restoration or plant phenotypes - Google Patents

Abstract

The present invention provides a process for the restoration of a plant phenotype that is altered due to a first transgene which when expressed inhibits expression of an endogenous plant gene, the process comprising introducing into said plant, or progeny thereof, a second transgene which encodes a protein or polypeptide that is capable of substituting the function of the protein or polypeptide product encoded by the said endogenous gene and wherein the nucleotide sequence identity of the tran-scripts encoded by the second transgene and the first transgene is less than 90 %.

Description

~ 094/09143 21 ~ 61 1 3 PCT/EP93/02875 Genetic noderation or restoration of plant phenoLy,~

FT~Tn OF THE lN V~N~l~10~
The present invention relates to genetically transformed plantsO methods for obt~ining genetically transformed plants and recombinant DNA for use therein. The invention further relates to a method for restoring a plant phenotype previous-ly altered due to the expression of a transgene in thatplant.
BACKGROUND ART
T~e European Patent Application 344 029 A2 describes a method for restoring male-fertility in plants that are male-sterile due to the expression of a first transgene encoA;ng R~rn~ in the tapetal cell layer of said plants, which method comprises the introduction into the same plant of a secon~ transgene encoding Barstar which is expressed at least in all those cells wherein the first transgene is expressed.
In the Barnase/Barstar syste~ for altering and restoring plant phenotype the first transgene, the Barnase gene is believed to interfere with a large number of endogenous gene products in a non-specific way, rather than by interaction with a preselected endogenous gene product. The restoration of male-fertility i8 based on a direct interaction of Barstar with Barnase. In general terms, fertility restoration accor-ding to this system is based on direct interaction of the restoration gene product with the sterility gene product in the plant cell. This is one of the best described phenotype restoration systems known in the art. However, a drawback of the R~rn~-~e/Barstar system is that its application is limited to phenotypes which allow disruption of cell structures by cell death. Phenotypes that re~uire more subtle modification of plant ce_l functioning, such as alteration of flower colour, fruit ripening, and the like, are outside the scope of this system.
Many systems for altering plant phenotypes are based on inhibition of endogenous plant genes. Examples thereof include but are not limited to disease-resistance, flower co-lour, fruit-ripening, male-sterility, and the like. It is an object of the invention to provide a phenotype restoration or moderation system that can be used when plant phenotypes have , .'. ~

2 i 4 61 l 3 - 2 - PCT/EP93/02 ~

been altered due to the expression of a transgene capable of inhibiting expression of a particular endogenous gene.

SUMMARY OF THE INVENTION
The present invention provides a process for the resto-ration of a plant phenotype that is altered due to a first transgene which when expressed inhibits expression of an endogenous plant gene, by introducing into said plant, or progeny thereof, a second transgene which when expressed is capable of neutralising or partially neutralizing the effect caused by the first transgene, whereby said second transgene is expressed at least in those cells involved in the altered phenotype. Preferred in a process according to the invention is a second transgene which encodes a protein or polypeptide gene product that is capable of substituting the function of the protein or polypeptide product encoded by the said endogenous gene and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 90%, preferably less than 80%, yet more preferably said second transgene encodes a protein or polypeptide gene product that is not identical in amino acid se~uence to the endogenous gene product and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 75%. According to a special preferred embodiment said second transgene is obtainable from a different plant species.
The invention further provides a process for the resto-ration of fertility in a plant that is male-sterile due to a first transgene which when expressed inhibits expression of an endogenous plant gene required for pollen development or functioning, by introducing into said plant a second trans-gene capable of neutralising the effect caused by the first transgene, whereby said second transgene is expressed in all cells in which the first transgene is expressed. Preferred in a process according to the invention said second transgene encodes a protein or polypeptide gene product that is capable of substit~ting the function of the protein or polypeptide product encoded by the said endogenous gene and wherein the nucleotide sequence identity of the transcripts encoded by 21~ ~113 PCT/EP93/02875 the second transgene and the first transgene is less than 90~ , preferably less than 80%, more preferably said second transgene encodes a protein or polypeptide gene product that is not identical in its amino acid sequence to the endogenous gene product and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 75%.
According to a special preferred embodiment said second transgene is obtainable from a different plant species.
According to a special embodiment the process according to the invention said first transgene is an antisense gene which when expressed inhibits expression of an endogenous flavonoid biosynthesis gene and said second transgene encodes a flavonoid biosynthesis enzyme capable of substituting the function of the corresponding flavonoid biosynthesis enzyme encoded by the said endogenous gene. Preferred according to this embodiment is a first transgene which is an antisense gene inhibiting expression of an endogenous chalcone synthase gene and said second transgene encodes a chalcone synthase - 20 capable of substituting the function of the chalcone synthase encoded by the said endogenous gene. Especially preferred first transgenes and second transgenes for the restoration or moderation of male-fertility are those obtA;n~hle from table 1 in this specification.
Preferred in a process according to the invention is the process wherein said second transgene is introduced intc the progeny of said plant by cross-pollination of a parent of said plant with pollen comprising said second transgene.
The invention further provides a process for ob~;n;ng fertile hybrid seed of a self-fertilizing plant species, comprising the steps of cross-pollinating a plant S which is male-sterile due to a transgene which when expressed inhibits expression of an endogenous gene required for normal pollen development or functioning, with a plant R which is male-fertile and comprises a transgene that encodes a protein orpolypeptide product capable of substituting the function of the protein or polypeptide product encoded by the said endogenous gene. Preferred according to this process is a first transgene which is an antisense chalcone synthase gene, WO94/09143 214 61 1~ . PCT/EP93/02 ~

the endogenous gene is a chalcone synthase gene, and the second transgene encodes chalcone synthase, wherein the nucleic acid sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 90%, preferably less than 80%, more preferably less than 75%.
The invention also comprises fertile hybrid seed obtained by a process according to the invention, as well as plants obtained from fertile hybrid seed, as well as parts of the plants, such as a bulb, flower, fruit, leaf, pollen, root or root culture, seed, stalk, tuber or microtuber, and the like.
The invention further comprises plants, as well as parts thereof, which harbour a chimeric gene which when expressed produces a protein or polypeptide product capable of substi-tuting the function of a polypeptide or protein encoded by anendogenous gene of said plant, wherein the nucleotide sequence identity of the transcripts encoded by the transgene and the endogenous gene is less than 90~, preferably less than 80%, more preferably less than ?5%.
D:E:SCRIPTION OF THE FIGURES
Figure l. A representation of plasmid MIP289 harbouring an expression cassette with multiple cloning site, which can be suitably used to insert foreign genes and antisense genes for expression in anthers of plant cells; CHI PB: chalcone isomerase B promoter; NOS tail: transcription termination signal derived from the nopaline synthase gene of Aqrobacte-~ium.
Figure 2. Same plasmid as in figure l, wherein the expression cassette contains a hybrid promoter based on the 35S promoter of cauliflower mosaic virus, and a so-called anther box (for details of promoter, vide Van der Meer, et al, 1992, sub) Figure 3. Crossing scheme for obtaining fully male-fertile hybrid seed according to the invention; plant S(Ssrr): maternal male-sterile line keterozygous for the sterility gene which when expressed inhibits expression of an endogenous plant gene required for pollen development or functioning; plant R: pollinator line heterozygous for a ~ 094/09143 2 i 4 6113 PCT/EP93/0287~

restoration transgene capable of neutralising the effect caused by the first transgene.

Figure 4. Similar crossing as in Figure 3, except for the pollinator line which is homozygous for the restoration gene.
J

Figure 5. Binary vector pFBP125. This is a pBIN19 based vector with an insert comprising a chs gene from Arabido~sis thaliana between a hybrid promoter fragment comprising the CaMV 35S RNA promoter in which an anther-box (AB) has been inserted, and the nos-termination region of Agrobacterium tumefaciens.

Figure 6. Binary vector pFBP130. This ~s a pBIN19 based vector with an insert comprising an chs gen~ from Arabido~sis thaliana between a promoter fragment of the chs-A gene of Petunia hYbrida and the n~-termination region of Agrobacte-rium tumefaciens.

Figure 7. Southern analysis of plant DNA of several petunia lines cont~ining: (a) petunia anti-sense chs con-struct (T29), (b) ArabidoPsis sense chs gene construct (-T36004), (c) both constructs (a) and (b) (T38002 and T38007) and wild-type (W115) p~obed with 32P-labelled Arabido~sis chs DNA (o/n exposure -80 degr. Celsius). The Arabidopsis chs genes are clearly visible in T38002 (several strong bands), T38007 (several strong bands) and T36004 (one strong upper band), whereas there is only slight cross-hybridizatiQn with the endogenous petunia chs genes or antisense petunia chs genes (faint bands in the lanes of T38002, T38007, T29 and W115 and the antisense gene in T29).
Figure 8. Northern analysis of messenger RNA of the same plants as in Fig. 7, including now T38005. Probed with petunia chs DNA; 6 days ~poCllre -80 degr. Celsius). The chs mRNA are clearly visible in the lanes of T36004 and W11~ as WO94/09143 21~ 6 I 13 6 PCT/EP93/02 -expected. In none of the antisense plant lines (T29, T38002, T38005, T38007) could a petunia mRNA be detected, as could have been expected as well.

Figure 9. Northern analysis as in Figure 8, except that the blot was probed with Arabidopsis chs DNA, o/n exposure at -80 degrees Celsius. At o/n exposure the Arabidopsis chs MRNA is only detected in the lane of T36004. However, upon gross overexposure some very faint bands could be detected in the lanes of the double transgenic lines T38002, T38005 and T38007.

DETAILED DESCRIPTION
The instant invention will be illustrated by outlining in more detail the findings that are obtained when performing experiments aimed at restoration of male-fertility in plants that were made male-sterile by thé expression in the tapetal cell layers of a chalcone synthase transgene which was placed in the reverse orientation with respect to the promoter. The details of the gene constructs and the male-sterile plants obtained therewith are described in Van der Meer et al., (1992, The Plant Cell 4, 253-262).
It was shown that expression of an antisense CHS gene in the anthers of transgenic plants caused inhibition of normal pollen functioning as a result of which the plant were unable to self-pollinate. The transgenic male-sterile plants were found to be entirely female-fertile and could be made to set seed by cross-pollination with a male-fertile pollinator line. It was concluded that the antisense chs plants can be suitably used for the production of hybrid crops.
In the experiments that underlie the present invention a male-sterile Petunia hybrida plant S which is transgenic for an antisense CHS gene from Petunia hybrida under the control of regulatory sequences that provide for expression of the transgene in anthers of the plants, is cross-fertilised with a Petunia hybrida plant R that contains a transgene obtaina-ble from the chs gene of ArabidoPsis thaliana which is under the control of regulatory sequences that pro~ide for ex-pression of the transgene in anthers of the plants.

~ 094/09143 214 6113 - PCT/EP93/02875 Of the pollinator plants R, harbouring only the transge-ne from Arabidopsis thaliana the majority is not male-sterile as mi~ht have bee expected from the finding that transgenes can inhibit the expression of resident genes enco~;ng homolo-gous gene products. This so-called co-suppressive effect has been established for a number of genes including a chs transgene obtainable from Petunia hYbrida and re-introduced into petunia plants (Napoli C. et al., 1990, The Plant Cell 2, 279-289; Van der Meer I. et al., 1992, Plant Cell 4, 253-262). It has also been disclosed that expression of a chs transgene placed in the sense direction under the control of its promoter gîves rise to male-sterile plants, just as expression of an antisense chs gene does, provided expression of the transgenes occurs at least in the tapetal cell layer of the anthers of the plants (PCT/NL92/00075, which is herewith incorporated by reference in this specification, with the proviso ~hat the definitions in that application do not apply to the description of this invention and the claims attached thereto at present or after A~çn~ment).
The finding that the introduction of a divergent chs gene, such as the one from Arabido~sis, does not markedly inhibit the production of chalcone synthase in the transgenic plants indicates, that significant co-suppressive effects are absent if a transgene is selected that encodes a transcript that is sufficiently divergent from the endogenous gene transcript.
The crossing of male-sterile plant S, which is heterozy-gous for the sterility gene (Ssrr) with plant R, homozygous for the restoration gene (ssRR) yields hybrid seed SR of which 50% contains in addition to the endogenous chs gene and the ArabidoPsis chs gene in the sense orientation, the antisense chs gene from Petunia hybrida. Contrary to expecta-tion, it will be found, that a percentage of the progeny plants grown from the hybrid seed (50% SsRr; 50% ssRr) harbouring both the transgenes is again capable of self-fertilization in spite of the fact that about 50% also inherited the sterility gene.
To establish the nature of the restored phenotype a transcript specific primer extension experiment is carried WO94/09143 PCT/EP93/02 ~
21~611~ 8 -out on CDNA obtained from young anthers. Attempts to visuali-ze radioactive extension products corresponding to the first (pet~nia chs) transgene transcript fails, which can be expected in view cf the restored phenotype. Applying equal radio-illumination times it is also impossible to detect the presence of the endogenous chs gene transcript, whereas an extension product of about l.4 kb obtained with the primers represented as SEQIDNO: l and SEQIDNO: 2 corresponding to Arabido~sis chs transgene transcript can be clearly detected under these conditions. The corollary of these experiments is that the endogenous gene transcript and the almost identical petunia transgene transcript interact, presumably by basepai-ring, as a consequence whereof these transcripts are not expressed and probably degraded in the plant nucleus. It is presumably due to the nucleic acid sequence divergence of the Arabidopsis transgene with respect to both the endogenous petunia gene, as well as the petunia transgene, that the former does not interact with any of the transcripts encoded by the latter two genes. The nucleic acid sequences of the Arabido~sis transgene and the Petunia gene transcripts differ at least 30% in the protein encoding region, presumably even more if the non-translated regions of the transcript are taken into account. Hence, the nucleic acid divergence of the transcript is deemed responsible for its translatability in the plant cell, thereby producing a fully active chalcone synthase which substitutes the endogenous chalcone synthase.
As a result male-fertility is restored in a percentage of the progeny plants despite the fact that about 50% thereof contain the sterility transgene.
Apparently, the high degree of nucleic acid sequence identity of the first (petunia) chs transgene antisense transcript and the endogenous (petunia) chs transcript favours the interaction of these molecules, probably causing them to be degraded, while the second chs transgene trans-cript from Arabido~sis thaliana which is at the most 75%
identical on the nucleic acid level (see Table l), is pro-duced in sufficient quantities to be translated into a fully functional (heterologous) chalcone synthase capable of restoring the plant's altered phenotype. We therefore main-VO94/09143 21~ 3 9 PCT/EP93/0287~

tain that the restoration of the male-fertility phenotype is due to complementation on the enzyme level.
This is believed to be the first observation of partial phenotype restoration, or phenotype moderation, in plants, wherein the production of an endogenous protein product is blocked and wherein the function of that protein product is substituted by a protein product similar (not necessarily identical) on the amino acid level, but encoded by a nucleotide sequence which is different on the nucleic acid level. This finding may have interesting applications in the genetic modification, restoration, or moderation of plant phenotypes, in and ou~side the area of hybrid seed produc-tion. For example, it is now feasible to silence endogenous enzymes, and substitute such enzymes by enzymes with differ-ent properties, such as a different substrate specificity,mode of regulation, and the like. Such substitutions may bring about subtle, yet interesting, changes in the biochemi-cal pathway in which the endogenous enzyme is involved.
The various aspects of the invention are outlined in more detail below.
The invention can be worked with any phenotype alterati-on system that involves an inhibitory gene of the antisense type, such as described in EP 240 208 A2, directed against an endogenous gene. Evenly so, it can be worked with an inhibi-tory gene of the sense type, which work by the as yet notfully understood ~h~ m referred to as co-suppression, disclosed in Napoli et al., l990, supra. Examples of such phenotypes include, but are not limited to disease-resis-tance, drought-resistance, flower colour, fruit ripening, and the like.
The restoration gene must encode a transcript that is sufficiently divergent from both the endogenous gene trans-cript as well as the inhibitory transgene transcript and yet encodes a protein or polypeptide capable of substituting the function of the endogenous gene product. Phenotype resto-ration can be made absolute. Alternatively, phenotype resto-ration may be made not absolute; in this case it is preferred to speak of partial phenotype restoration or 'phenotype moderation'. If absolute phenotype restoration is aimed at, W094/09143 214 6113 lo - PCT/EP93/02 ~

the di~ergence of the transcript must diverge preferably by more than 20%, that is the nucleic acid identity of the restoration transcript with either the inhibitory transgene transcript or the endogenous gene transcript does not exceed 80~, preferably it does not exceed 75%. Depending on the level of moderation desired, optimal moderation can be achieved by making transgenes with different levels of divergence and selecting the desired phenotype. In case phenotype restoration is not required to be absolute, or desired to be not absolute, divergence of the restoration transgene transcript should not exceed 20%, preferably it should not exceed 10%. The latter is referred to as phenotype moderation.
Likewise, phenotype alteration systems that involve inhibitory genes of the ribozyme type directed as sequence specific endo-ribonucleases against an endogenous gene trans-cript, as disclosed in US Patent 4,987,071, may be restored with a transgene according to the invention, with the proviso that the restoration gene encodes a transcript that is lacking the recognition and/or cleavage consensus of the ribozyme. Phenotype moderation should be possible using this kind of inhibitory transgenes as well, although manipulating the recognition and cleavage sequence of the restoration gene to affect its affinity for the ribozyme may require some trial and error.

The choice of the restoration gene As a rule the restoration gene must not give rise to a transcript that is identical to the endogenous gene trans-cript. Preferably, the restoration gene transcribed region isas much divergent from the transcribed region of the endoge-nous gene as possible, while the protein product encoded by said transcript is identical, or almost identical. It is well known in the art that each amino acid can be encoded by a more than one codon; this fact, referred to as the degeneracy of the genetic code, stems from the fact that there are about 20 different amino acids, which are encoded by triplets of four different bases, yielding a total of 64 possible codons.
Three codons comprise stop signals for translation, so that ~ 094/09143 2 14 6 113 PCT/EP93/02875 in actual fact 61 codon specify about 20 amino acids. Roughly spoken, every third base may be changed in a coding region without affecting the amino acid sequence of the protein.
This means that the transcribed region of a restoration gene can at least diverge 33% from the endogenous gene. But, since a gene transcript generally comprises non-translated regions flanking the coding region on both sides, even further nucleic acid divergence may be achieved in order to avoid interaction of the restoration gene transcript with the endogenous gene transcript or the first transgene transcript.
Furthermore, still greater divergence may be achieved if one takes into account the fact that two prôteins may differ in their amino acid sequence, while ret~;n;ng their physiolo-gical activity in the plant cell. Although it is not esta-blished to what extent this may be, it may be assumed thatproteins which have conservative amino acid replacements in 10% of their amino acids, will st^ill be capable of performing their physiological role. Alto~ether, it will be clear to ~omeone skilled in the art that a restoration gene according to the present invention need not be more identical to its endogenous counterpart than about 40-50% on the nucleic acid level.
Some aspects of the invention will be further illustra-ted with male-sterility as exemplifying phenotype.
Obtention of a male-sterile maternal line S
Any male-sterile plant phenotype that is due to expres-sion of an inhibitory gene of one of the types mentioned in the preceding paragraphs can be restored by a restoration gene according to the invention.
Typical examples of how genes can be identified that are essential for pollen development or pollen functioning is given inter alia in KO89/10396 and KO90/08828. Once such genes are isolated they can be expressed or overexpressed in the sense or antisense orientation in those cells required for pollen development or functioning. In order to achieve expression in those cells that are necessary for pollen development, genes are placed under the control of promoters that are expressed in stamen cells (including filaments and W O 94/09143 2 ~ 4 6113 - 12 - PC~r/EP93/02 ~

anthers), or more specifically in anthers, or even more specifically in tapetal cell layers thereof. A distinction should be made to sterility genes that are disruptive to general plant cell functioning or viability on the one hand, and genes that disrupt plant metabolism to the extent that it disrupt pollen development or functioning without drastically affecting plant viability on the other hand. The antisense chalcone synthase gene is one of the latter category; conse-quently, it is not necessary for the latter type sterility gene to be expressed exclusively in stamen cells through the use of stamen-specific promoters. Sterility genes of the former type, i.e. the general plant cell disrupters, must not be effective inside plant structures essential for survival of the plant. Methods for isolating promoters that provide for proper expression patterns of these genes are also described in both WO89/10396 and WO90/08828, which are herewith deemed incorporated by reference.
For reasons of illustration the maternal male-sterile line is represented as being heterozygous for the sterility gene. However, it will be clear that fully fertile hybrid seed can be obtained also if the maternal line is homozygous for the sterility gene. International Patent Application PCT/NL92/00075, discloses a method for obtaining homozygous male-sterile plants, by selfing male-sterile plants harbou-ring one copy of an antisense chs gene, whereby the pollenthat are arrested in their development are made to germinate on pistils in the presence of flavonoids. The seed obtained from this selfing can be grown into homozygous male-sterile maternal plant lines, which can optionally be propagated n ~itro first, and then used as such in hybrid seed production by cross-pollination with a pollinator line, which may be heterozygous or homozygous for the restoration gene according to the invention.

Plant transformation Introduction of sterility genes, herbicide resistance genes or restoration genes into plants, is achieved by a any one of the following techniques, the choice of which is not critical to the present invention.

~ 0 94/09143 21~ 6 1 13 ~ PC~r/EP93/02875 Generally, useful methods are the calcium/polyethylene glycol method for protoplasts (Krens, F.A. et al., 1982, Nature 296, 72-74; Negrutiu I. et al, June 1987, Plant Mol.
Biol. 8, 363-373), electroporation of protoplasts (Shillito R.D. et al., 1985 Bio/Technol. 3, 1099-1102), microinjection into plant material (Crossway A. et ~1., 1986, Mol. Gen.
Genet. 202, 179-185), (DNA or RNA-coated) particle bombard-ment of various plant material (Klein T.M. et al., 1987, Nature 327, 70), infection with viruses and the like.
Preferred according to the invention is the use of Aqrobacterium-mediated DNA transfer. Especially preferred is the use of the so-called binary vector technology as disclo-sed in EP-A 120 516 and U.S. Patent 4,940,838).
Subsequently, receptive plant cells or are selected for the presence of one or more markers which are encoded by plant expressible genes co-transferred with the plant expres-sible gene according to the invention, whereafter the trans-formed material is regenerated into a whole plant. Alternati-ve~r, pollen cells are transformed, for instance by coated-particle acceleration, and used to pollinate receptiveplants.
Although considered somewhat more recalcitrant towards genetic transformation, monocotyledonous plants are amenable to transformation and fertile transgenic plants can be regenerated from transformed cells. Presently, preferred methods for transformation of monocots are microprojectile bombardment of explants or suspension cells, and direct DNA
uptake or electrop. ~ation (Shimamoto, et al, 1989, Nature 338, 274-276). Transgenic maize plants have been obtained by introducing the strePtomYces hvqroscoPicus bar-gene, which encodes phosphinothricin acetyltransferase (an enzyme which inactivates the herbicide phosphinothricin), into embryogenic cells of a maize suspension culture by microprojectile bombardment (Gordon-Kamm et al, 1990, Plant Cell, 2, 603-618). The introduction of genetic material into aleuroneprotoplasts of other monocot crops such as wheat and barley has been reported (Lee, 1989, Plant Mol. Biol. 13, 21-30~.
Wheat plants have been regenerated from embryogenic suspensi-on culture by selecting only the aged compact and nodular WO94/09143 214 ~113 - 14 - PCT/EP93/02 ~

embryogenic callus tissues for the establishment of the embryogenic suspension cultures (Vasil I., et al, 1990, Bio/Technol. 8, 429-434). Herbicide resistant fertile wheat plants were obtained by microprojectile bombardment of regenerable embryogenic callus (Vasil V. et al, 1992, Bi-o/technol. 10, 667-674).The combination with transformation -systems for these crops enables the application of the present invention to monocots.
Monocotyledonous plants, including commercially impor-tant crops such as corn are also ~en~hle to DNA transfer byAqrobacterium strains (Gould J, Michael D, Hasegawa O, Ulian EC, Peterson G, Smith RH, (1991) Plant. Physiol. 95, 426-434).

~arker qenes Suitable marker genes that can be used to select or screen for transformed cells, can be selected from any one of the following non-limitative list: neomycin phosphotransphe-rase genes conferring resistance to kanamycin (EP-B 131 623), the hygromycin resistance gene (EP 186 425 A2) the Glutathione-S-transferase gene from rat liver conferring resistance to glutathione derived herbicides (EP-A 256 223), glutamine synthetase conferring upon overexpression resistan-ce to glutamine synthetase inhibitors such as phosphinothri-cin (W087/OS327), the acetyl transferase gene from Streptomy-ces viridochromogenes conferring resistance to the selective agent phosphinothricin (EP-A 275 957), the gene encoding a 5-enolch;k;~te-3-phosphate synthase (EPSPS) conferring tole-rance to N-phosphonomethylglycine, the kar gene conferring resistance against Bialaphos (e.q. W091/02071), and the like.
The actual choice of the marker is not crucial as long as it is functional (i.e. selective) in combination with the plant cells of choice.
The marker gene and the gene of interest do not necessa-rily have to be linked, since co-transformation of unlinked genes (U.S. Patent 4,399,216) is also an efficient process in plant transformation.

Gene exPression ~094/09143 2 1 4 6 113 15 - PCT/EP93/02875 The expression pattern reauired for the restoration gene depends on the expression pattern of the inhibitory transge-ne. The latter in its turn is dependent on the phenotype alteration aimed at. Thus, for modifying the fruit ripening phenotype in a plant, an inhibitory gene bringing about said alteration must at least be expressed in the fruits of said plant. Restoration or moderation can be achieved by an expression pattern that comprises at least the expression pattern of the inhibitory transgene.
Multiple transqenic Plants To obtain transgenic plants harbouring more than one gene a number of alternatives are available, the actual choice of which is not material to the present invention:
~. the use of one recombinant polynucleotide, ~ ~ a plasmid, with a number of modified genes physically coupled to one selection marker gene.
B. Cross-pollination of transgenic plants which are already capable of expressing one or more chimeric genes coupled to a gene encoding a selection marker, with pollen from a trans-genic plant which contains one or more gene constructions coupled to another selection marker. Afterwards the seed, which is obtained by this crossing, is selected on the basis of the presence of the two markers. The plants obtained from the selected seeds can afterwards be used for further cros-sing.
C. The use of a number of various recombinant polynucleoti-des, e.a. plasmids, each having one or more chimeric genes and one other selection marker. If the frequency of cotrans-formation is high, then selection on the basis of only onemarker is sufficient. In other cases, the selection on the basis of more than one marker is preferred.
D. Consecutive transformations of transgenic plants with new, additional genes and selection marker genes.
E. Combinations of the above mentioned strategies.
The actual strategy is not critical with respect to the described invention.

selection of hybrid seed WO94/09143 2 ~ 4 6 113 PCT/EP93/02 ~

It is known in the art that, the need to separate hybrid seed from non-hybrid seed can be avoided if the self-pollina-tors can be destroyed, for example by using an antibiotic, preferably a herbi~ide. This requires that the maternal male-sterile line is resistant to this antibiotic or herbicide dueto the presence of transgene coding therefor.
The herbicide resistance gene may be introduced into the maternal line simultaneously with the sterility gene accor-ding to the invention by genetic transformation with a multi-gene construct. However, the herbicide resistance gene may beintroduced into the maternal line after the introduction of the sterility gene.
It may be advantageous to introduce the herbicide resistance trait into the plant intended to use as maternal parent line prior to the introduction of the sterility gene.
This simplifies the creation of plants that are homozygous for the herbicide resistance phenotype which may be advanta-geous. Then, plants provided subsequently with the sterility gene, may be cross-pollinated with a pollinator plant contai-ning a restoration gene according to the invention. Suitable herbicides can be selected from any one listed under the heading marker genes.

Advantaqes and industrial aPPlication The process according to the invention is particularly useful for the production of hybrid progeny that is fully male-fertile.
In a conventional process of producing hybrids from self-fertilising crops a transgenic (heterozygous) nuclear male-sterile plant line S (Ssrr) may be crossed with a male-fertile plant line R (ssrr) to yield hybrids that are 50~
fertile (ssrr) and 50% sterile (Ssrr). Consequently, if ~-uch hybrid crops were grown in the field directly, 50% of the acreage would consist of plants that must be cross-fertilised in order to set seed, which may have significant yield reducing effects for those crops that rely on the setting of fruit or seed for their commercial value. Examples of such crops include but are not limited to cereals and oil seed rape.

~0 94/09143 21~ 6113 PC~r/EP93/02875 Thus, the present invention is especially suitable for the hybridization of naturally self-fertilizing crops by crossing a maternal line which is male-sterile due to the expression of a first transgene capable of inhibiting expres-sion of an endogenous plant gene essential to normal pollenfunctioning, and a pollinator line containing a second transgene capable of neutralising the effect caused by the first transgene. Although 50% of the hybrid progeny is heterozygous for the sterility gene, the presence of the restoration or moderation gene ensures fertility of the progeny that is closer to that of the wild type lines.
The specific advantages of this hybridization system reside in the fact that it can be used in combination with any sterility system that makes use of transgenes inhibitory to endogenous genes. As a consequence the phenotype can be determined predominantly by the nature of the gene product, rather than the specificity of the expression pattern.
All references cited in this specification are indicati-ve of the level of skill in the art to which the invention pertains. All publications, whether patents or otherwise, referred to previously or later in this specification are herein incorporated by reference as if each of them was individually incorporated by reference.
The Examples given below are just given for purpo-ses of illustration and do not intend in any way to limit the scope of the invention.

Construction of a chiPB/as-chs and a chalcone isomerase B
promoter chs qene construct (chiPB/chs-At~
The chiPB/as-chs construct comprises a chs cDNA fragment from Petunia hybrida fused in the antisense orientation to a chalcone isomerase B promoter fragment. The chiPB/chs-At construct comprises a chs cDNA fragment from Arabidopsis thaliana fused in the sense orientation to a chalcone isomerase B promoter fragment.
A 1.7 kb promoter fragment from the anther-specific chiPB promoter (Tunen, A.J. Van., Mur, L.A., Brouns, G.A., Rienstra, J.D., Koes, R.E. and Mol, J.N.M., 1990, The Plant W094/09143 214 6113 1 PCT/EP93/02 ~

Cell 2, 393-401) and a 0.2 kb NOS tail isolated from plasmid pBI101.1 (Jefferson, R.A., Kavanagh, T.A., and Bevan, M.W.
(1987). EMBO J. 6, 3901-3907) are cloned into the plasmid pUC19 (Messing, J., 1978, Recombinant DNA Technical Bulletin NIH Publication No. 79-99, 2, 43-48) yielding the recombinant plasmid MIP289 (Figure 1).
A 1.4 kb BamHI chs fragment is isolated from plasmid pTS21 (Van der Meer et al., 1992, su~ra) and cloned into plasmid MIP289 digested with BamHI. A clone with the chs fragment in an antisense orientation is selected on the basis of the asymmetric SstI restriction enzyme site. Subseguently, this fragment is subcloned as a HindIII/EcoRI fragment into the binary vector Binl9 (Bevan, M. (1984) Nucl. Acid Res. l~, 8711-8712) yielding plasmid pAS8.
To isolate a full size Arabidopsis chs cDNA, single stranded cDNA is synthesized on 10 ~g RNA isolated from young Arabido~sis thaliana ecotype Landsberg erecta flower buds, by priming with an 17-mer oligo-dT primer (Maniatis, T., Fritsc-h, E.F., and Sambrook, J. (1982). Molecular Cloning: A
Laboratory Manual (Cold Spring Harbour, NY: Cold S~ring Harbour Laboratory). A set of two additional primers based on (Feinbaum, R.L., and Ausubel, F.M. (1988). Mol. Cel. Biol. 8, 1985-1992) with the sequence based on the 5' region (primer I
2 SEQIDNO: l; GCGGATCCGTATACTATA~GGTGATGG) and 3' region (primer II = SEQIDNO: 2; GAGGATCCTTAGAGAGGAACGCTGTGCAAGAC) of the Arabidopsis chs gene are used for the initial polymerase chain reaction (PCR) analysis. The PCR reaction is performed in 100 ~l PCR buffer (10 mM Tris, pH 8.3, 50mM KC1, 2.5 mM
MgC12) containing 50 pmole primers, and 200 ~M of each deoxy-nucleotide triphosphate. Amplification involved 30 cycles ofa st~n~rd cycle for homologous primers. Amplified CDNA is fractionated on a 1% agarose gel and z 1.4 kb band is iso-lated and subcloned as a BamHI fragment (sites present in the 5' and 3' primers) in pAS8 after digestion with BamHI to remove the petunia chs CDNA. The orientation and proper cloning of the ArabidoPsis chs CDNA into PAS8/BamHI is checked by a detailed restriction enzyme analysis and sequence analysis; the correct plasmid is called pAS9.

~ 094/09143 2 14 6 113 PCT/EP93/0287S

Example 2 Transformation of tobacco plants ~ he plasmids pAS8 and pAS9 are transferred from E. coli JM83 (Messing et al, 1978, su~ra) to Agrobacterium tumefa-ciens strain LBA 4404 (Hoekema A. et al., lg83, Nature 303:179-180) by triparental mating (Rogers, S.G., and Fraley, R.T., 1985, Science 227, 1229-1231), using a strain contai-ning plasmid pRK2013 (Ditta et al., 1980, Proc. Nat. Ac. Sci.
USA, 12, 7347-7351). Transformed tobacco plants are obtained by the st~n~rd leaf-disc transformation method (Horsch et al., 1985, Science 227, 1229-1231). After cultivation with the A. tumefaciens strains harbouring either pAS8 or pAS9, the tobacco leaf discs are grown on MS plates conta;ning 3 ~g/ml kinetin, 500 ~g carbenicillin and 200 ~g kanamycin.
Plants obtained are checked for transformation on th~ basis of resistance for kanamycin and by Southern blot ana~ysis using an ~ fragment as a probe. After shoot and root induction plants are put on soil and transferred to the greenhouse. Plants are grown under in the greenhouse at 21-C
at a 16 hours light, 8 hours dark regime.

ExamPle 3 Analysis of transqenic plants expressinq the antisense chs construct 25Transgenic tobacco plants cont~in;ng the ~h;m~ric pAS8 gene construct (Petunia antisense chs) are investigated for fertility by self-pollination. At least one plant is almost completely sterile and shows a seed set of less than 1% in selfings. Furthermore the pollen grains of this plant are morphologically aberrant, as was also published by Van der Meer et al. (1991) and are not able to germinate in an ia vitro germination assay. This plant is designated Sl and contains only one copy of construct pAS8 in its genome.

35Exam~le 4 ~nalYsis of transqenic ~lants expressinq the chimeric Arabi-do~sis chs construct From a number of 15 transgenic tobacco plants containing plasmid pAS9, one plant expressing the Arabidopsis chs cDNA

WO94/09143 2 1 ~ 6 1 1~ - 20 - PCT/EP93/02 ~

in young anthers is selected by RNAse protection experiments using RNA isolated from young anthers. This plant is designa-ted Rl.

Exam~le 5 Crossing of Sl and Rl restores fertility A cross is made between Sl (genotype Ssrr) and Rl (genotype ssRr) and the offspring of this cross is grown to mature plants. Based on their genotype four classes of plants can be distinguished: Ssrr, SsRr, ssRr, and ssrr (see also Figure 2). It can be observed that plants containing the restoration gene, i.e. the ArabidoPsis chs gene (SsRr) are able to set seed after self-pollination despite the presence of a sterility gene (Ss). Light-microscopical analysis shows that these plants have pollen that are morphologically normal whereas Ssrr plants have aberrant pollen. All plants contai-ning both the sterility gene construct pAS8 and the restora-tion gene construct pAS9 show restoration of fertility as can be demonstrated by self-pollination experiments. In a control cross between Sl and an untransformed tobacco plant only 50%
of the offspring is able to set seed after self-pollination as can be expected on the basis of the fact that Sl has a copy of construct pAS8 integrated in its genome.

The following table provides data about chalcone synthase genes from various plant species and the nucleic acid identity of the amino acid coding regions: reference sequence is Petunia hybrida U30 chalcone synthase gene. Best match is given at a minimum sequence of lO00 bp.

TABLE l Comparison of NA sequence identity of chs genes 35 sou-ce qene desiqnation identity (%) P. nYbr_da V30 chs lO0 P. lY~r_~a c~s~ 98 P. ny~r ~a c~sJ 82 P. nY~r_~a c~sH 79 40 P. nYJr ~a c~sD 77 P. ~y~r La c~sF 7~
P. ~Y~r ~a c~sG 76 ~094/09143 214 ~ PCT/EP93/02875 . esculentum TCHSl 83 L. es-ulentum TCHS2 83 P. sa- vum PSCHSl 76 P. sa_ vum PSCHS2 74 5 P. sa__vum PSCHS3 73 Soybean CHS gene 3 74 G. max CHS gene 2 73 Parsley CHSl 72 M. incana CHSy 71 10 A. thaliana Atchs 71 Mustard SasCHS3 72 Mustard SasCHSsg 70 Mustard SasCHS1 69 Antirrhinum maius AmCHS 74 15 Pinus s~lvestris PsCHSs 70 Hordeum vulqare X58339 68 Zea maYs Zmc2cs 67 Zea mays Zmwpcs 67 Boldface: gene fragments that are used as sterility and restoration gene respectively, in this disclosure.
Other suitable combinations of sterility genes and restora-tion genes can be selected from this table.

~XAMPLE 7 Partial fertility restoration in male-sterile ~lants bY re-transforming male-sterile ~lants with a diverqent restoration gene construct Petunia W115 plants were transformed with a sterility gene construct comprising the promoter region of the petunia chs gene linked to the coding region of the petunia chs gene in antisense orientation. This gene construct, designated VIP176 (Krol A.R. van der et al., 1990, Plant Molecular Biology 14, 457-466) was used to transform the petunia line W115 and a self-sterile plant was selected and designated T17002. This self-sterile flavonoid depleted plant, T17002, was cross-pollinated with a W115 plant and among the progeny a plant was selected, which was kanamycin sensitive but still self-sterile and depleted for flavonoids; this plant was designated T29. This sterile, kanamycin sensitive plant, T29, - was used for a second transformation with pFBP125 (PCaMv35sAB/
CHSAt, yielding the 39000 plants~ not discussed further) or pFBP130 (PCHS pet/CHSAt rendering the 38000 plants, see below).
This approach was successful as 7 transgenic 38000 plants were obtained which contain both the sterility gene WO94/09143 214 6 1~ PCT/EP93/02 construct (chs-antisense from petunia) as well as the restor-ation construct (sense-chs from Arabidopsis). Of these plants 5 had flavonol production in the corolla; 2 out these 5 plants were male-fertile (inter alia T38005).
In order to obtain data about the functionality of the Arabidopsis CHS-enzyme in petunia plants, W115 plants were transformed with ~grobacterium strains harbouring pFBP125 (yielding the 36000 plants, see below). Of 15 transformed plants, 4 plants over-produced flavonols as compared to wild-type (W115) (inter alia T36004, see Table 2).
Plant lines were tested for the presence of the con-structs by Southern analysis. Expression of the genes was verified by Northern analysis.
Table 2 summarizes the results for 6 petunia lines: from top to bottom are given Southern data, obtained by probing with petunia chs probes and Arabidopsis chs probes: Northern data, obtained by probing with both aforementioned probes, corolla pigmentation (flavonol staining); and fertility determination. The genetic backgrounds of the petunia lines are as follows: W115 - wild-type petunia plants (non-trans-genic); T29 - Pc~M~5sAB-antisense petunia chs (transgenic for sterility gene); T38002, T38005, T38002 - Pchs-antisense petunia çhs + Pchs-A.thaliana chs (transgenic for sterility gene and restoration gene); T36004 ~ PCaM~5SAB A- thalian (transgenic for the restoration gene only).
As indicated in the table lines W115, which is 100%
fertile, and T29, which is entirely unable to self-pollinate, performed as expected (see PCT/NL92/00075). The double-transgenic lines T38002, T38005 and T38007, which contain both the sterility gene and the fertility gene, had only a partially restored fertility; for T38005 seed-set was about 10-20% of the wild-type W115. These data correspond well with the presence of only slight amounts of flavonols (see below~.
Moreover, the presence of flavonols was dependent on the presence of the ArabidoPsis chs gene, as was confirmed by Southern data using the Arabidopsis chs PCR fragment as a probe. The ArabidoPsis chs probe was only weakly capable of cross-hybridizing with the petunia chs gene and vlce versa (Fig. 7).

~146113 The Northern data on mRNA of corolla's corresponded with the Southern data, except that the Arabi~-~sis chs-messenger RNA of plant lines T38002, T38005 and T38007, when probed with the Arabidopsis chs-probe, could only be detected after gross over-exposure; this is probably due to weak expression of the Pchs-Arabidopsis chs gene construct in corolla's. The Northern data for lines T38002, T38005 and T38007 seem in accordance with production of low amounts of flavonoles in these lines, which, in turn, might explain the fact that the sterility was restored only partially (only 10-20~ seed set for T38005 as compared to W115). In order to restore fertil-ity it is necessary that the restoration gene construct (such as in pFBP125 and pFBP130) is expressed in either the male reproductive organs or the female reproductive organs or in both. Although expression of the restoration gene in corol-la's provides an initial indication of fertility restoration it will be necessary to establish expression of the Arabido~-sis gene in either of the reproductive organs. We anticipate that the T38005 plant expresses the ArabidoPsis gene in either of the reproductive organs, and Northern analysis is in progress to confirm this.
The proper functioning of the Arabidopsis CHS-enzyme was established by comparing flavonol production in W115 with T36004, which contains, in addition to its endogenous chs-gene, copies of the Arabidopsis chs-gene. As is indicated, corolla's of T36004 produced far more flavonols (+++ - dark orange, after staining for flavonols) than W115 corolla's (+
= pale orange). As expected, male-sterile line T29 did not produce detectable amounts of flavonols (- = purely white corolla's), whereas the corolla's of T38002, T38005, which was partially fertility-restored, and T38007 produced slight amounts of flavonols in the corolla (+/- = beige or very pale orange).
The high flavonol levels observed in corolla's of T36004 correspond well with the Northern data obtained for that plant line, indicating abundant levels of the Arabidopsis chs-messenger RNA in these lines (see Fig. 9). It is, there-fore, clear that the ArabidoPsis CHS-enzyme is fully func-tional in petunia plants and, in principle, capable of WO94/09143 2 14 6113 PCT/EP93/02 ~

substituting the function of endogenous CHS.

DePosited microorganisms On October 14, 1993, two E. coli JM101 strains, one harbouring pFBP125, and one harbouring pFBP130 have been deposited at the Centraal Bureau voor Schimmelcultures, Baarn, The Netherlands, under accession number CBS 543.93 and CBS 544.93, respectively.

~0 94/09143 2 14 6 1 13 25 - PCI/EP93/02875 o a~

O UJ 1) ~
O ~ J

a) U
+ + ~ ,~ U ~ U~

~ X

~ O O
O U~
,v '~ t ' V o V ~1 + ~ I ~ U h~ CU¦ ~C E3' t U ~ Ci ~ C ~C~ U ~

~ CC o~C" ~ e~C ~, (~ ) UJ a~
+ + I I I I I ~ U~ ~ ~ ~ ) "C

C~ Ul 1 In u ,,,~ P-~ ~

a) d~ 0 h 3 s ;~
+ I + I + ~ .A ~

:~ - . t t ,~ ~ h _ UJ ~ S 4 ~ ~

- S -- h '~ r "~ ~J
.. . _ ~ S ~ ~ ~ r ~ ~ ~ ~ U~ 0 . = ~ . , ~ 0 ~ U R

~) ~ r-l t IV ~ ~ U. 0 ~ ~ Cl ::" .::',.. ,'~ U ~ * ~; ~ K ~ ~1 ~ .C 1~ 0 c ~ 4 t .~

W O 94/09143 P ~ /EP93/02 ~

SEQUENOE IISTING
(1) GENERAL INFORM~IION:
(i) APPIIC~NT:
(A) NAME: MOGæN Tnt~rn~tiona1 N.V.
(B) ST~EET: Einsteinweg 97 (C) crrY: T~
(D) SI~E: Zuid-~
(E) C~UN~ : me N~thPrl ;~
(F) P0SIAL CODE (ZIP): NLr2333 CB
(G) T~TFP~NE: (0)31.71.258282 (H) ~ FAX: (0)31.71.221471 (ii) TTTIE OF INVENTION: Genetic Re~ lin~ of P1ant Phenntypes (iii) N~MBER OF SEQUENCES: 2 (iv) CX/E~IrER REanPBLE FORM:
(A) MEDIUM IYPE: F1cppy disk (B) CXPD~IrER: IBM PC cr~tihl~
(C) OPERAIING SYSTEM: PC-DOS/MS-D06 (D) SOFIW~RE: PatentIn ~PlPA~P #1.0, Version #1.25 (EPO) (2) INFoRMaIICN FOR SEQ ID NO: 1:
(i) SEQUEN OE CH~RA~~ sllCS:
(A) IENGTff: 28 base pairs (B) T~PE: ~lr-lP;c acid (C) SIl~n~NE95: single (D) T0~0L0GY: linear ~ ;ul~: T~PE: cDNA to mRN~
(iii) H~u~ l~L: YES
(~i) ORIGIN~L SoURCE:
(A) CEr~ 3M: Ar~h;~ç;ç th~ nA
(B) STRAIN: l~ y erecta (F) ~l~l.c~u~ T~PE: Flawer budç
(X) F~JRT,Tt'~TT(l~ l~p~CN:
(A) A~ ~S: Feinbaum, R L
All~lhPl, F M
(B) TITIE: ~11~ r ;~kional re~ t;~n of the Ar~h;~r~;ç
th~ na rh~l~nne synth~ce gene (C) JOURN~L: MD1. Cell. Biol.
(D) VOLUME: 8 (F) PAGES: 1985-1992 (G) n~IE: 1988 (xi) SE ~ CE DES~x~ uN: SEQ ID NO: 1:

(2) INECRMPIION FOR SEQ ID NO: 2:

VO 94/09143 21~ 6 113 PCI/EP93/0287S

(i) SEX~ENOE C~RA~
(A) LENG~I: 32 hase pairs (B) T~E: ~ rlP;r- acid (C) s~s c; ~1 (D) IOPOIOGY: liT~ear (ii) ~TF~T~ q~YPE: clNA to ~NA
(iii) ~Y~ L: YES
(vi) ORIG~L Sa~ROE:
(A) OR~N~: Ar~h;~ th;~l ;~rla (B) STR~ y erecta (F) ~l~u~; T~: flaWP~ ~c (A) A~ ;: Feiriba~n, R L
A l~lhPl, F M
(B) mlE: Tr~lll~ riptional regulation of the Ar~h;tlff ~ic thA~ a rh~lr~e synthase (C) J wFa~L: Mol. Cell. Biol.
(D) VolDME 8 (F) PAGES: 1985-1992 (G) n~IE: 1g88 (xi) ~u~ DES~xl~ll~: SEQ ID NO: 2:
Ga~XPDrOCTT pr.Ar.~rr.~C ~ A~G AC 32

Claims (23)

1. A process for the restoration of a plant phenotype that is altered due to a first transgene which when expressed inhibits expression of an endogenous plant gene, by introducing into said plant, or progeny theroef, a second transgene which when expressed is capable of neutralising or partially neutralizing the effect caused by the first transgene, whereby said second transgene is expressed at least in those cells involved in the altered phenotype.

2. A process according to claim 1, wherein said second transgene encodes a protein or polypeptide gene product that is capable of substi-tuting the function of the protein or polypeptide product encoded by the said endogenous gene and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 90%.

3. A process according to claim 2, wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 80%.

4. A process according to claim 3, wherein the said second transgene encodes a protein or polypeptide gene product that is not identical in amino acid sequence to the endogenous gene product and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 75%.

5. A process according to any one of the claims 1 - 4, wherein said second transgene is obtainable from a different plant species.

6. A process for the restoration of fertility in a plant that is male-sterile due to a first transgene which when expressed inhibits expression of an endogenous plant gene required for pollen development or functio-ning, by introducing into said plant a second transgene capable of neutralising the effect caused by the first transgene, whereby said second transgene is expressed in all cells in which the first transgene is expressed.

7. A process according to claim 6, wherein said second transgene encodes a protein or polypeptide gene product that is capable of substi-tuting the function of the protein or polypeptide product encoding by the said endogenous gene and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 90%.

8. A process according to claim 7, wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 80%.

9. A process according to claim 8, wherein the said second transgene encodes a protein or polypeptide gene product that is not identical in its amino acid sequence to the endogenous gene product and wherein the nucleotide sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 75%.

10. A process according to claim 6 - 9, wherein said second transgene is obtainable from a different plant species.

11. A process according to any one of the claims 6 to 10, wherein said first transgene is an antisense gene which when expressed inhibits expression of an endogenous flavonoid biosynthesis gene and said second transgene encodes a flavonoid biosynthesis enzyme capable of substituting the function of the corresponding flavonoid biosynthesis enzyme encoded by the said endogenous gene.

12. A process according to claim 11, wherein said first transgene is an antisense gene inhibiting expression of an endogenous chalcone synthase gene and said second transgene encodes a chalcone synthase capable of substituting the function of the chalcone synthase encoded by the said endogenous gene.

13. A process according to any one of the claims 7 - 12, wherein said first and said second transgene are selected from the group consisting of the chalcone synthase genes obtainable from table 1 in this specifica-tion.

14. A process according to any one of the claims 1 to 13, wherein said second transgene is introduced into the progeny of said plant by cross-pollination of a parent of said plant with pollen comprising said second transgene.

15. A process for obtaining fertile hybrid seed of a self-fertilizing plant species, comprising the steps of cross-pollinating a plant A which is male-sterile due to a a transgene which when expressed inhibits expres-sion of an endogenous gene required for normal pollen development or functioning, with a plant B which is male-fertile and comprises a trans-gene that encodes a protein or polypeptide product capable of substitu-ting the function of the protein or polypeptide product encoded by the said endogenous gene.

16. The process of claim 15, wherein the first transgene is an antisen-se chalcone synthase gene, the endogenous gene is a chalcone synthase gene, and the second transgene encodes chalcone synthase, wherein the nucleic acid sequence identity of the transcripts encoded by the second transgene and the first transgene is less than 90%.

17. The process of claim 16, wherein the nucleic acid sequence identity of the transcripts encoded by the second transgene and the first transge-ne is less than 80%.

18. The process of claim 17, wherein the nucleic acid sequence identity of the transcripts encoded by the second transgene and the first transge-ne is less than 75%.

19. Fertile hybrid seed obtained by the process of claim 15.

20. Plants obtained from seed of claim 19, as well as parts of the plants, such as a bulb, flower, fruit, leaf, pollen, root or root culture, seed, stalk, tuber or microtuber, and the like.

21. A plant, as well as parts thereof, which harbour a chimeric gene which when expressed produces a protein or polypeptide product capable of substituting the function of a polypeptide or protein encoded by an endogenous gene of said plant, wherein the nucleotide sequence identity of the transcripts encoded by the transgene and the endogenous gene is less than 90%.

22. The plant and plant parts of claim 21, wherein the nucleotide sequence identity of the transcripts encoded by the transgene and the endogenous gene is less than 80%.

23. The plant and plant parts of claim 22, wherein the nucleotide sequence identity of the transcripts encoded by by the transgene and the endogenous gene is less than 75%.