CA2290883C - Method for enhancement of naturally occurring cytoplasmic male sterility and for restoration of male fertility and uses thereof in hybrid crop production - Google Patents

Abstract

The present invention relates to methods for enhancement of naturally occurring cytoplasmic male sterility and for restoration of male fertility and uses thereof in hybrid crop production. There is also disclosed a method for restoration of male fertility to cytoplasmic male sterile plants; which comprises the steps of: a) introducing into the nucleus of a plant cell a gene construct essentially consisting of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an edited form of a normal mitochondrial gene that is co-transcribed with an unusual CMS-associated mitochondrial gene; b) selecting for plan cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant.

Description

METHOD FOR ENHANCEMENT OF NATURALLY OCCURRING
CYTOPLASMIC MALE STERILITY AND FOR RESTORATION OF MALE
FERTILITY AND USES THEREOF IN HYBRID CROP PRODUCTION
BACKGROUND OF THE INVENTION
(a) Field of the Invention The invention relates to methods for enhancement of naturally occurring cytoplasmic male sterility and for restoration of male fertility and uses thereof in hybrid crop production.
(b) Description of Prior Art For many crops, hybrids, formed by crossing two or more different strains, provide higher yields than do the parental strains themselves. For example, in canola (Brassica napus, campestris), Canada's most important crop, manually produced hybrids have yields that can be up to 50% greater than those of the paren-tal lines. This phenomenon, termed hybrid vigor, has been most successfully applied in maize (Zea mays).
Hybrid maize constitutes approximately 90% of the North American crop.
To produce hybrid seed on a commercial scale, it is necessary to prevent self-pollination of the seed parent of the hybrid cross. This is a relatively sim-ple matter in maize, because the male or pollen-produc-ing organs (the stamens) are located in a different part of the plant than the female, or seed-bearing organs (the carpels), and the stamens, which collec-tively form a structure called the tassel, can easily be removed manually from the large numbers of plants used in a seed production operation. By contrast, in most crop species, such as canola, the stamens and car-pels are located in the same floral structure, and manual emasculation of large numbers of plants is not possible. Hence, seed producers require alternative methods of pollination control for producing hybrids in these crops. A review of the technology developed for this purpose can be found in The PBI Bulletin article of Arnison P. (Arnison P. et al., The PBI Bulletin, January 1997, 1-11).
One alternative method is to use chemicals, called gametocides, that specifically kill pollen.
These chemicals are generally expensive and are often only partially effective, For most crops, especially those like canola that flower over an extended period, this type of pollination control is not cost-effective.
Nearly all hybrid seed production systems therefore rely on genetic control of pollination. Genetic polli-nation control mechanisms fall into three categories:
self-incompatibility, "molecular hybridization"
methods, and cytoplasmic male sterility.
Self incompatibility (SI) results from the capacity of some plants to reject their own pollen.
Plants expressing SI can be used as the female line in hybrid production, but, because these plants normally reject their own pollen, it is usually difficult, and cost-ineffective, to propagate or maintain such lines.
In addition, self-incompatibility is not found in most plant species. Molecular hybridization methods are the most recently developed. They rely on genetically engineered male sterility, caused by the specific expression of toxic proteins in pollen forming cells.
Such introduced toxin genes act as dominant male ste-rility genes and can be used to generate female lines for hybrid seed production. As with SI, the propaga-tion of such female lines is not straightforward and involves loss of plant material. These usefulness of these methods is still unclear, and at present they are used in the pro3uction of only a few hybrid varieties and these are not widely grown.

Cytoplasmic male sterility (CMS), is a wide-spread and classic non-Mendelian trait_ CMS plants are incapable of self-pollination and hence when a CMS line is planted alongside a male-fertile line, all the seed that forms on the sterile plants will be a hybrid of the two parents. Unlike most traits, CMS is maternally transmitted, i.e., it is passed on to offspring only through the seed parent. This property results from the fact that the gene or genes that determine CMS are located on mitochondrial DNA (mtDNA); unlike most genes, which reside in nuclear DNA, genes in mtDNA are transmitted solely through the female in most plant species. As a result of this property of maternal transmission, it is possible to easily propagate female CMS lines, by pollinating with a male fertile line "maintainer" line, that is identical to the CMS line in its nuclear genes, but which is male fertile because in lack the CMS-causing mtDNA. However, again as a result of the maternal transmission of CMS, hybrid plants pro-duced using female CMS lines would also be male sterile because they would carry the male-sterility conferring mtDNA. This is problematic for seed crops such as maize and canola, which require pollen production for the formation seed, the harvested product. Fortu-nately, in many crop species specific dominant nuclear genes termed restorers of fertility (Rf) have been identified that can suppress the male-sterile phenotype and "restore" fertility to Fl hybrids. The components of a CMS system therefore consist of the CMS line, that contains the male sterile (S) cytoplasm (or mtDNA) and is homozygous for the recessive or maintainer allele of the restorer gene, the maintainer line, that contains a fertile or normal mtDNA (F) and but is isogenic with the CMS line at nuclear genetic loci, and the a restorer line, that usually contains the male sterile ---- --- ---- - ------mtDNA but is homozygous for the dominant nuclear Rf gene. The use of these components in a hybrid seed production scheme is illustrated in Fig. 1.
To produce a diverse set of hybrids using CMS
adequate numbers of restorer lines, that contain Rf genes, as well as "maintainer" lines, that are steril-ized by the CMS mtDNA, must be available. The develop-ment of these lines through conventional genetics is a slow process that minimally requires several years of effort. For example, to develop a new restorer line it is necessary to first cross the recipient line with an existing restorer line, to introduce the Rf allele. A
series of backcrosses are then required to recover the genotype of the recipient line. Even after many gen-erations of backcrosses some donor DNA linked to the Rf gene will remain, a phenomenon termed linkage drag;
this donor DNA may carry deleterious traits and compro-mise the quality of the recipient strain.
Two CMS systems are proving to have some limited use in hybrid canola seed production. The Polima or pol cytoplasm is capable of conferring male sterility on many canola cultivars, and effective restorer lines, possessing dominant alleles at a single nuclear restorer locus, have been identified. The cost-effectiveness of hybrid production using this system is limited by the fact that po1-induced male sterility tends to be incomplete or "leaky", especially when the female line is exposed to warmer growing conditions.
As a result, pol hybrid seed may be contaminated with CMS seeds that result from self-pollination of the female line. As these CMS plants are male sterile and do not spontaneously set seed, their presence can considerably lower the overall yield the "hybrid"
mixture. A second CMS system is based on the use of a hybrid cytoplasm in which the male sterility determinant is derived from a radish cytoplasm termed Ogura or ogu. Male sterility conferred by the ogu cytoplasm is complete. Restorer lines for this system, which were developed by introgressing a single radish restorer gene (which we designate here as Rfo) into Brassica napus, have recently become available, but the development of restorer lines is hampered by "linkage drag" between the restorer gene and genes that have a negative effect on seed quality.
It would be highly desirable to be provided with methods for enhancement of naturally occurring cytoplasmic male sterility and for restoration of male fertility and uses thereof in hybrid crop production.
STJHMARY OF THE INVENTION
One aim of the present invention is to provide methods for enhancement of naturally occurring cytoplasmic male sterility and for restoration of male fertility and uses thereof in hybrid crop production.
In accordance with the present invention there is provided a method for enhancing naturally occurring cytoplasmic male sterility in plants; which comprises the steps of:
a) introducing into the nucleus of a plant cell a gene construct essentially consisting of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with, at least one of, an unedited form of atp6 gene and an orf224 gene of Brassica napus mitochondria;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant.
In accordance with the present invention there is also provided a method for restoration of male fertility to cytoplasmic male sterile plants; which comprises the steps of:
a) introducing into the nucleus of a plant cell a gene construct essentially consisting of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an edited form of a normal mitochondrial gene that is co-transcribed with an unusual CMS-associated mitochondrial gene;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant.
The preferred plant is Brassica napus.
Preferably, step b) is effected using a plant transformation vector, such as pRD400.
In accordance with the present invention there is also provided a method for restoration of male fertility to polima cytoplasmic male sterile B. napus;
which comprises the steps of:
a) introducing into the nucleus of a B. napus plant cell a gene construct essentially consisting of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an edited form of an atp6 gene of B. napus mitochondria;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration.of selected plant cells to produce a mature plant.
Our results have implications with respect the practical implementation of po1 CMS for the production of hybrid canola and other Brassica crops. The intro-duction, into po1 CMS plants, of genetic constructs that allow the targeting of the product of the edited form of the Brassica atp6 gene to the mitochondria rep-resents a new type of process by which male fertility restored plants can be recovered; this should allow the production of new pol CMS restorer lines in a single step, through plant transformation. This would repre-sent a considerable savings in time and money over the conventional plant breeding methods for generating such lines, which is at present, the only way in which these lines can be generated. In addition, the introduction of constructs that allow the targeting of the products of the unedited form of the atp6 gene or or.f224 into the mitochondria to produce plants with enhanced male sterility represents a process through which new po1 CMS lines with enhanced maintenance of male sterility could potentially be produced; as mentioned above, the leakiness of pol male sterility is the major impediment to its mor.e widespread use in hybrid canola production and there is no effective means of addressing this.

BRIEF DESCRIPTION OF THE DRAWINGS
Fig. +1 illustrates the use of cytoplasmic male sterility (CMS) in hybrid seed production;
Figs. 2A-2J illustrate the flowers of fertile, pol CMS and transgenic B. napus cv. Westar plants;
Fig. 2Ashows complete flowers of pol CMS Westar ( left ) and pol CMS Westar transformed with the mas2'/A9-A6e construct (note the increased size of the petals in the transgenic plant);
Fig. 2B shows complete flowers of partially male sterile transgenic plant obtained by introducing AP3/A9-A6u into Westar (nap) (left) and male fertile Westar (nap) (note the decreased pigmentation of the petals of the transgenic flower);
Fig. 2C shows flowers with petals and sepals removed of pol CMS Westar (left) and po1 CMS Westar transformed with the mas2'/A9-A6e construct; note the increase size of the stamens and the more highly devel-oped anthers in the transgenic plant;
Fig. 2D shows flowers with petals and sepals removed of partially male sterile transgenic plant obtained by introducing AP3/A9-A6u into Westar (nap) (left) and male fertile Westar (nap) (note the reduc-tion in anther size in the transgenic plant);
Fig. 2E shows complete flowers of (left to right) pol CMS Westar, male sterile transgenic plant 40-8 obtained by introducing the AP3/A9-ORF construct into male fertile Westar (nap), and male fertile Westar (nap);
Fig. 2F shows flower of the transgenic plant 40-8 with the petals and sepals removed; note that the four inner stamens have been transformed into carpels;
Fig. 2G shows complete flowers of the partially fertile transgenic plant 40-10 obtained by introducing the AP3/C4-ORF construct into male fertile Westar (nap) (left); a flower of the recipient strain Westar (nap) is shown on the right (Note the reduction in petal size of the transgenic plant);
Fig. 2H shows flowers with petals and sepals removed from transgenic plant 40-10 and male fertile Westar (nap);
Fig. 21 shows complete flowers of pol CMS Westar (left) and a partially fertile transgenic plant obtained by introducing the AP3/A9-A6e construct into pol CMS Westar (note the increased petal size of the transgenic plant);

_ 9 _ Fig. 2J shows flowers with petals and sepals removed of pol CMS Westar (left) and a partially fer-tile transgenic plant obtained by introducing the AP3/A9-A6e construct into po1 CMS Westar (note the increased size of the stamens and anthers in the trans-genic flower);
Figs. 3A-3D illustrate phenotypic changes accompanying mas2'-driven expression of A9-A6e in pol CMS Westar; A: Transgenic plant inflorescence; B: pol CMS Westar inflorescence; C: entire transgenic plant;
D: pol CMS Westar plant (Note the increased size of the inflorescence and vegetative internodes in the trans-genic plant);
Fig. 4 illustrates the use of AP3/A9-A6e construct as a dominant fertility restorer gene in B.
napus; and Fig. 5 illustrates the use of AP3/C4-ORF
construct to enhance pol cytoplasmic male sterility in B. napus.
DETAILED DESCRIPTION OF THE INVENTION
This invention provides a method for the devel-opment of maintainer and restorer lines through genetic engineering. The invention specifically relates to the pol_ CMS system of canola (Brassica napus, campestris).
This invention is based on research into the molecularbasis of cytoplasmic male sterility (CMS) in canola (Brassica napus L.) that had been conducted in my laboratory between 1987 and 1993. Analysis of Bras-sica mitochondrial genome organization and expression had shown that the polima or pol form of B. napus CMS

is correlated with expression of a mitochondrial gene region that contains a chimeric gene, orf224, that is co-transcribed with a normal mitochondrial gene, atp6 (Singh, M. P. and Brown, G.G. (1991). The Plant Cell 3, 1349-1362; Bonen, L. and Brown, G.G. (1993). Can. J.
Bot. 71, 645-660). We hypothesized that pol CMS is due to a mitochondrial dysfunction that results either from (i) the presence of the ORF224 protein in the mitochondrial membrane, or (ii) from a possible deficiency in the ATP6 protein that might be caused by interference of orf224 in atp6 translation.
The proteins of the mitochondrion have a dual genetic origin: some of these proteins, perhaps 10% or so, including those that specify CMS, are encoded in mtDNA; others are encoded in nuclear DNA. Proteins encoded in mtDNA are synthesized on mitochondrial ribo-somes, and remain within the mitochondrion. Nuclear-encoded mitochondrial proteins are synthesized as pre-cursors that usually contain an additional stretch of amino acids at the N-terminus called the transit pep-tide or presequence. This transit peptide serves to direct the protein to the mitochondrion; it is removed by a protease once the protein is inside the mitochon-drion. If the DNA sequence for a protein normally encoded in mtDNA is fused to DNA sequence of mitochon-drial transit peptide, and the resulting construct is introduced into and expressed in the nucleus, the resulting precursor protein can be imported into the mitochondrion and the final product can function in the same manner as the mitochondrially made form.
We reasoned that if hypothesis (i) was correct, then we might be able to confer male sterility on male fertile plants by expressing DNA constructs that would allow a nuclear-encoded, cytoplasmically translated ORF224 protein to be targeted to mitochondria. Simi-larly, we reasoned that if hypothesis (ii) was correct, we might be able to restore male fertility to pol CMS
plants by expressing DNA constructs that would permit a nuclear-encoded, cytoplasmically translated form of the ATP6 protein to be targeted to mitochondria. If the constructs altered male sterility/fertility, then the orf224 transgenes could be viewed as synthetic main-tainers of male sterility, while the atp6 transgenes could be viewed as synthetic restorers of male fertil-ity. A further complication to these hypotheses came from the observations that both orf224 and atp6 tran-scripts are, like most plant mitochondrial transcripts edited (Stahl R. et al. (1994) Nucleic Acids Res 22, 2109-2113). Plant mitochondrial transcript editing involves the conversion of specific C residues in an mRNA molecule to U residues which alters the amino acid sequence of the encoded protein (the edited mRNA codes for the functional form of the protein). Both the atp6 and orf224 transcripts are edited at a single site, resulting, in each case, in a single change in amino acid sequence. Since expression, in the nucleus, of a DNA construct that allowed targeting of an unedited wheat ATP9 protein to mitochondria conferred male ste-rility on transgenic tobacco plants (Hernould, M. et al. (1993). Proc. Natl. Acad. Sci. U.S.A. 90, 2370-2374), it seemed possible that expression/targeting of an unedited Brassica ATP6 protein could also cause male sterility and act as a synthetic maintainer gene.
Synthetic restorers of this type could, in principle, be used to create new restorer lines by genetic transformation. This would circumvent a lengthy and costly plant breeding exercise by which these lines are currently developed. The availability of synthetic maintainer genes might allow the genera-tion of lines through plant transformation that could completely maintain pol sterility. Such lines cur-rently do not exist, and this technology would there-fore effectively remove the final barrier to yield enhancement by pol based canola hybrids. Finally, it is possible, even likely, that analogous strategies could be used to generate synthetic restorer and main-tainer lines for crops other than canola and hence that a global strategy for genetically engineered mainte-nance and restoration of CMS could be developed.
Evidence that nuclear-encoded mitochondrial targeted forms of the ATP6 and ORF 224 proteins can act as syn-thetic restorer and maintainer genes, respectively To begin to address these possibilities we fused DNA corresponding to the edited (A6e) and unedited (A6u) forms of the atp6 coding sequence to DNA coding for either of two mitochondrial targeting presequences, yeast COX4 (C4) or Neurospora ATP9 (A9). Similar constructs were prepared by fusing the same presequences to the unedited form of orf224 (ORFu).
These sequences were joined, at their 5' end, to the mas2' promoter (van der Leegt-Plegt, L.M. et al.
(1992). Plant Cell Reports 11, 20-24) and at their 3' end to a polyadenylation signal for the Cauliflower Mosaic Virus_ 35S RNA. The resulting constructs were introduced into B. napus cotyledons by Agrobacterium-mediated transformation (Moloney, M. et al. (1989).
Plant Cell Rep. 8, 238-242). We were - not able to regenerate plants into which the ORFu constructs had been introduced. Three of the transgenic plants we recovered after introducing the A9/A6u construct into fertile B. napus cv Westar were partially male sterile, however, while the single plant that we recovered after introducing the A9/A6e construct into sterile pol CMS
B. napus was partially male fertile. This suggested that the construct expressing the unedited form of atp6 was acting as a sterility inducing, or maintainer gene, while the edited form was acting as a pol fertility restorer. Unfortunately, we were unable to recover seed from any of the plants expressing either construct, and so it was not possible to show that the changes in fertility/sterility could be genetically transmitted to subsequent plant generations or to confirm that the phenotypes resulted from transgene expression.
More recent research focused on the assembly of constructs in which a developmentally regulated pro-moter, derived from the APETALA3 (AP3) gene of Arabidopsis thaliana (Jack, T. et al. (1994) Cell 76, 703-716), was used to drive expression of atp6 and orf224 constructs. AP3 was chosen as a tissue specific promoter because it is expressed very early in the development of the stamen (as well as the petal).
Previous data have indicated that the arrest in pollen development associated with the po1 CMS is pre[neiotic and hence occurs early in stamen development; we therefore felt that to offset or mimic the pol CMS
phenotype it would be necessary to express the con-structs early in stamen development. Subsequent con-structs that have been completed or are currently being assembled include one in which the "constitutive" mas2`
promoter has been used for expression. All the con-structs were inserted into the binary transformation vector pRD400, which has been optimized for efficient transformation in Brassica, and introduced into plants through the cotyledon transformation procedure. A
summary of the constructs we have assembled, and the phenotypes of the transgenic plants we have recovered are indicated in Table 1, below.

Table 1 Genetic constructs and corresponding phenotypes of transgenic plants Constructl Introduced KanR Confirmed Phenotype(s) of confirmed transgenic into plants transgenic2 plants recovered AP3/C4-ORF3 fertile 17 12 Variable male sterility ranging from near Wastar complete sterility (2 plants) to complete fer-(nap) tility (most plants); some plants show altera-tions in floral organ number, morphoiogy &
identity AP3/A9-ORF " 21 7 Variable male sterility ranging from near complete sterility (1 plant) to complete fertil-ity (most plants); some plants show altera-tions in floral organ number, morphology &
identity AP3/A9-A6e male sterile 9 7 Ail plants partially male fertile; stamen and Westar petal size intemnediate between fertite (po1) Westar (nap) and Westar (pol) CMS
plants AP31A9-A6u fertile 20 9 Most plants male fertile with some flowers Westar partially sterile; some plants display reduced (nap) - stamens; pale yellow petals; changes in whorl identity and/or organ number.

mas2w- male sterile 22 16 Most plants partially male fertile; stamen and A6e Westar petal size intermediate between fertile (po1) Westar (nap) and Westar (po1) CMS
plants; some plants show changes in floral whori identity; elongated inflorescence and vegetative intemodes.

AP3/GUS Westar 13 9 Identical to Westar (nap); GUS expression (nap) observed at base of petals and stamens, as expected 1 Construct elements presented in sequence: promoter/presequence-coding sequence; AP3, Arabi dopsl s APETELA3 promoter, C4, yeast COX4 presequence; A9, Neurospora ATP9 presequence; ORF, edited orf224 coding sequence; A6e, edited a tp 6 coding sequence; A6u, unedited a tp 6 coding sequence. -2 Confirmed by Southern blot analysis using a probe derived from the NPTII
gene.
3 Epitope-tagged edited a tp 6 coding sequence fused to the A9 presequence.

The data of Table 2 indicates of the frequencies with which various constructs induced changes in male sterility or fertility in transgenic plants. Some of the changes observed in male fertility/sterility and floral structure in the transgenic plants are shown in Figure 2.
The flowers shown on the left side of Fig. 2A
and the right side of Fig. 2B are from pol CMS Westar and fertile Westar (nap) plants, respectively. The Figures beneath each of these (2C and 2D, respectively) show flowers from the same plants with the petals and sepals removed. As is evident in these figures, the stamens and petals of pol CMS plants are considerably smaller than those of Westar (nap) flowers. Under the controlled growing conditions we have employed (20 , 16h day, 15 , 8h night), the stamens that form on flow-ers of young po1 CMS plants are considerably less than half the height of the pistil, do not develop fully formed anthers and shed only small amounts of pollen.
Table 2 Numbers of transgenic plants with modified male fertility/sterility Construct Recipient plant Fertiiel Semi-fertile2 Semi-sterile3 Sterile4 AP3/A9-ORF Westar (nap) 6 3 1 AP3/C4-ORF Westar (nap) 5 5 2 AP3/A9-A6e po1 CMS 5 2 Westar AP3/A9-A6u Westar (nap) 9 6 mas2'/A9- po1 CMS 1 8 10 A6e Westar 1lndistinguishable from Westar (nap) 2Reduced pollen release in comparison to Westar (nap); filled seed pods obtained upon selfing 3Minimal pollen release; poor seed set upon selflng 4Few or no seeds obtained upon selfing; several mas2'/A9-A6e plants produced significant pollen but failed to set seed and hence appear to be female and male sterile.

As indicated in Tables 1 and 2, in about half of the transgenic plants we recovered, AP3-driven expression of the edited orf224 coding sequence, fused to either targeting presequence, proved capable of reducing male fertility and inducing changes in floral morphology in Westar (nap) plants. The degree of ste-rility, as assessed by the amount of pollen visible on the anthers of newly opened flowers, as well as the nature of other floral abnormalities, varied somewhat among different transgenic plants and occasionally, to a much smaller degree, among the different inflores-cence branches of individual plants. There was no apparent difference between the COX4 (C4) and ATP9 (A9) presequence constructs in either the conferred pheno-type or the frequency at which male sterility was induced. Some indication of the phenotypes obtained can be seen in Figs. 2 E-H, which show f lowers from the transgenic plants for which a change in pollen produc-tion was observed. In these plants the sizes of the petals and stamens as well as the amount of pollen pro-duced were reduced in comparison to Westar (nap), although not to the extent to which they are reduced in po1 CMS plants. The most extreme phenotype was observed with the transgenic plant 40-8, an A9-ORF
transformant (Figs. 2E & 2F): in these flowers, the four inner stamens were converted into carpel-like structures; the anthers that formed on the two outer stamens remained pale yellow and shed only very small amounts of pollen.
We assessed the capacity of individual plants to set seed upon self fertilization as a practical measure of their male/female fertility. To self plants we usually place a crossing bag over an inflorescence and gently tap it to promote pollination. In the case of normal male fertile plants, all the seed pods within the bag will be completely filled; pol CMS plants pro-duce no, or very few seeds under these conditions.
When we attempted to self transgenic plants expressing AP3/A9-ORF or AP3/C4-ORF constructs in this manner, we found that in one case no seeds (transgenic plant 40-8, Fig. 2), and, in several other cases, only a few seeds, were obtained. Those plants which displayed a marked reduction in seed set, but retained the capacity to produce some pollen were designated as "semi-sterile"
(Table 2). In total 3 of the 22 confirmed transgenic ORF plants were semi-sterile by this test.
Expression of the unedited form of the atp6 gene (A6u) fused to the ATP9 presequence, when driven by the AP3 promoter, also proved capable of reducing the fertility of Westar (nap) plants (Table 1). Slightly fewer than half of the recovered transgenic plants expressing this construct showed a reduction in male fertility (Table 2). In general, however, the reduction of male sterility, according to the capacity to set seed, was more pronounced than with the ORF con-structs; all the transgenic plants which showed a phe-notype were classified as semi-ster-ile. Most of the plants with reduced fertility also possessed small, pale yellow petals as well. The semi-sterile phenotype displayed by one of the AP3/A9-A6u plants is shown in Figs. 2B and D. We have also found that constructs of this type, as well as the ORF constructs and a third type of AP3/A9-A6 construct, AP3/A9-A6ep, in which the ATP6 coding sequence is innterrupted by a sequence encoding a small peptide, can enhance the sterility of po1 CMS plants. pol CMS plants normally produce a small amount of pollen; po1 CMS plants expressing AP3/A9-A6u shed no pollen.

As shown in Figs. 21 and J, AP3-driven expres-sion of the edited atp6 construct in pol CMS plants caused a change in phenotype consistent with the pre-diction that it acts as a synthetic restorer gene.
Flowers on these transgenic plants had petals and sta-mens intermediate in size between those of po1 CMS and fertile Westar (nap) plants. The anthers of these flowers developed small, yellow anthers that produced significant amounts of pollen, although not as much as male fertile Westar (nap) plants (Table 1). All the transgenic plants displayed an altered phenotype (Table 2), although the amount of pollen produced varied slightly from plant to plant. These plants were all capable of producing seed upon bagging of the inflorescence; two of the plants produced only small amounts of seed, while the five remaining plants produced filled pods without bagging or other type of intervention. These latter 5 plants can be therefore be regarded as being fertility restored in a full functional sense of the term.
We also assessed the capacity of the A9-A6e construct to induce male fertility in po1 CMS plants when expressed using the "constitutive" promoter mas2'.
Six of the transgenic plants we recovered possessed flowers that were virtually identical to those produced by AP3/A9-A6e plants (Fig. 1A & 1C). The remaining plants produced either smaller amounts or virtually no pollen. Only one of the pollen producing plants was capable of forming significant numbers of seeds without being manually pollinated which suggests that the female fertility of the mas2'/A9-A6e plants is, in gen-eral, reduced. This is consistent with previous results with constructs of this type in which we recov-ered plants with fertile flowers that are incapable of setting seeds. A second feature of the plants we recovered is also consistent with our preliminary results: all the plants expressing A9-A6e from the mas2` promoter showed changes in vegetative growth. In particular, both vegetative and inflorescence inter-nodes we longer than those of Westar (nap) or pol CMS
Westar plants (Fig. 3); in some transgenic plants leaf morphology was also affected. Similar changes were not observed in any of the plants in which employed the AP3 promoter for expression. Thus both the reduction in female fertility and other abnormalities result from the generalized expression of fusion protein from the mas2' promoter.
The observation of similar types of sterility modifications in multiple independently transformed plants provides very strong support for our initial hypotheses, namely: (i) that constructs which allow targeting of the ORF224 or unedited ATP6 polypeptides to the mitochondria can confer sterility on male fer-tile plants and hence serve as synthetic maintainer genes; and (ii) that constructs which allow targeting of the edited ATP6 polypeptide to the mitochondria can serve as synthetic restorer genes. The data also indi-cate that do avoid deleterious effects on vegetative growth and female fertility, it is necessary to express these constructs using a tissue specific or inducible promoter. The analysis of the R1 progeny of the RO
plant (a plant recovered from the transformation -regeneration protocol) A9-ORF expressing plant 40-8 (see above) has provided compelling evidence that the phenotypes observed are inheritable and due to the transgene. We raised 20 of these plants to maturity.
A range of floral phenotypes were evident, from the extreme abnormalities of the parental plant, where only two very sterile anthers form and the remaining four anthers are converted to pistil-like organs, to the partial male fertility observed in many of the RO
plants that express this construct. All the R1 progeny plants contained the transgene, suggesting that the original 40-8 RO plant may have had more than one copy of the gene. This would further indicate that the degree of sterility expressed is dependent on gene dosage. This gene dosage effect provides an additional means of modifying the degree of sterility expressed by the transgenic plants.
In summary, our experiments represent the first case in which male sterility has been induced through the targeting of the products of mitochondrial ORFs (orf224 and the unedited atp6) in a homologous host plant, and the first time that Brassica genes have been used for this purpose. Our results strongly also suggest that AP3 driven expression of the A9-A6e con-struct can serve as a fertility restorer gene for the po1 CMS. This is the first case we are aware of in which a natural form of CMS has been suppressed by a synthetic gene construct.
Our results are consistent with experiments conducted by other researchers, who have found the expression/targeting of the protein products of a bean CMS-associated ORF (He, S. et al. (1996). Proc. Natl.
Acad. Sci. U.S.A. 93, 11763-11768) or an unedited wheat ATP9 gene (Hernould, M. et al. (1993). Proc. Natl.
Acad. Sci. U.S.A. 90, 2370-2374), in a heterologous recipient plant, tobacco, can induce male sterility.
Interestingly, similar types of experiments with CMS-associated ORFs from maize and petunia have failed to result in the recovery of male sterile plants (Chaumont, F. et al. (1995). Proc. Natl. Acad. Sci. USA
92, 1167-1171; Wintz H. et al. (1995). Plant Mol Biol 28, 83-92). The frequency with which we recovered male sterile plants was higher than was observed for the bean ORF, and difficult to compare with the results using the unedited wheat ATP9 gene.
The present invention will be more readily un-derstood by referring to the following examples which are given to illustrate the invention rather than to limit its scope.

EXAMPLE I

Use of the AP3/A9-A6e construct as a dominant fertility restorer gene in B. napus In this example, outlined in Fig. 4, the A9/A6e construct is introduced by genetic transformation into the genome of a variety such as Westar which is normally sterilized by (pol) cytoplasm. The gene would be introduced into pol cytoplasm line, and the resulting transgenic plants recovered would be expected to be partially male fertile. This line is then self-fertilized to generate a partially male fertile restorer line that is homozygous for the A9/A6e transgene. This line could then be crossed with different po1 CMS lines to generate fertility restored Fl hybrids.

EXAMPLE II

Use of the AP3/C4-ORF construct to enhance pol cytoplasmic male sterility in B. napus The sterility induced by pol cytoplasm on most B.napus genotypes is incomplete, especially at higher temperatures. This residual pollen production results in a variable degree of self-fertilization in the CMS
line in a hybrid seed production operation. Plants raised from these seeds will not be hybrid and will have a reduced capacity themselves to set seed. These factors together can reduce, to a considerable extent, the yield of the "hybrid" seed batch. In this example, outlined in Fig. 5, the dominant sterility conferring properties of C4-ORF are employed to reduce or eliminate residual pollen production in the CMS line, and thereby enhance the percentage of hybrid seed recovered, and the hence yield of the hybrid seed batch. The C4-ORF construct is introduced by genetic transformation into a fully male fertile Westar (nap) line to produce a partially male fertile line that contains the C4-ORF transgene. This line is then self-fertilized to generate a partially male fertile maintainer line and is also crossed with a non-transformed line of the same nuclear genotype but with pol cytoplasm, to generate a completely sterile po1 CMS
line that is heterozygous for C4-ORF.
This is backcrossed to the C4-ORF partially male fertile maintainer line to generate a completely male sterile pol CMS line that is homozygous for C4-ORF.
This CMS line, when crossed to a pol restorer line, will produce a fertile Fl hybrid. The completely sterile pol CMS line is propagated by crossing it, as female, with the nap cytoplasm, partially male fertile transgenic C4-ORF maintainer line.
Although we have described a method for enhancing the sterility of a pol CMS plant using the AP3/C4-ORF construct, the AP3/A9-A6u and AP3/A9-A6ep constructs could also be used for this purpose, as the expression of all these constructs result in completely male sterile pol CMS transgenic plants.
While the invention has been described in con-nection with specific embodiments thereof, it will be understood that it is capable of further modifications and this application is intended to cover any varia-tions, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains and as may be applied to the essential features hereinbefore set forth, and as follows in the scope of the appended claims.

Claims (22)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method for enhancing naturally occurring polima-induced cytoplasmic male sterility in Brassica napus plants;
which comprises the steps of:
a) introducing into the nucleus of a B. napus plant cell a gene construct comprising a developmentally regulated promoter operably linked to a DNA sequence encoding a mitochondrial transit peptide fused upstream of and in frame with, at least one DNA sequence encoding an unedited form of a B. napus atp6 gene transcript, an edited form of a B. napus orf224 gene transcript, or both, wherein said promoter is expressed during stamen development;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature cytoplasmic male sterile plant.

2. The method of claim 1, wherein said promoter is AP3.

3. The method of claim 1 or 2, wherein step a is effected using a plant transformation vector.

4. A method for restoration of male fertility to polima-induced cytoplasmic male sterile Brassica napus plants; which comprises the steps of:
a) introducing into the nucleus of a B. napus polima-induced cytoplasmic male sterile plant cell a gene construct comprising, a developmentally regulated promoter operably linked to a DNA sequence encoding a mitochondrial transit peptide fused upstream of and in frame with a DNA sequence encoding an edited form of the B. napus mitochondrial atp6 gene transcript, wherein said promoter is expressed during stamen development;

b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature male fertile plant.

5. The method of claim 4, wherein said promoter is AP3.

6. The method of claim 4 or 5, wherein step a is effected using a plant transformation vector.

7. A method of reducing pollen production in plants, which comprises the steps of:
a) introducing into the nucleus of a plant cell a gene construct comprising a developmentally regulated promoter driving expression of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an edited form of the orf224 gene of Brassica napus mitochondria, wherein said promoter is expressed early in stamen development;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant with reduced pollen production.

8. The method of claim 7, wherein the plant is Brassica napus.

9. The method of claim 7 or 8, wherein step a) is effected using a plant transformation vector.

10. The method of any one of claims 7 to 9, wherein said promoter is AP3.

11. The method of any one of claims 7 to 10, wherein the plant is male fertile.

12. The method of any one of claims 7 to 10, wherein the plant is partially male sterile.

13. A method of reducing pollen production in partially male sterile Brassica plants, which comprises the steps of:
a) introducing into the nucleus of a partially male sterile Brassica plant cell a gene construct comprising a developmentally regulated promoter driving expression of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an unedited form of the atp6 gene of B. napus mitochondria, wherein said promoter is expressed early in stamen development;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant with reduced pollen production.

14. The method of claim 13, wherein the plant is Brassica napus.

15. The method of claim 13 or 14, wherein step a) is effected using a plant transformation vector.

16. The method of any one of claims 13 to 15, wherein said promoter is AP3.

17. A method of reducing pollen production in plants, which comprises the steps of:
a) introducing into the nucleus of a plant cell a gene construct comprising a developmentally regulated promoter driving expression of a sequence encoding a mitochondrial transit peptide fused upstream of and in frame with an edited form of the orf224 gene of Brassica napus mitochondria, wherein said promoter is expressed during any stage of stamen development;
b) selecting for plant cells that have acquired the gene construct in step a); and c) inducing regeneration of selected plant cells to produce a mature plant with reduced pollen production.

18. The method of claim 17, wherein said promoter is AP3.

19. The method of claim 17 or 18, wherein the plant is male fertile.

20. The method of claim 17 or 18, wherein the plant is partially male sterile.

21. The method of any one of claims 17 to 20, wherein the plant is a Brassica plant.

22. The method of claim 21, wherein the plant is a Brassica napus plant.