WO1990008831A1

WO1990008831A1 - Inhibition of plant cell respiration

Info

Publication number: WO1990008831A1
Application number: PCT/GB1990/000114
Authority: WO
Inventors: Ian George Bridges; Simon William Jonathan Bright; Andrew James Greenland; Wolfgang Walter Schuch; Graeme Alexander Reid
Original assignee: Imperial Chemical Industries Plc
Priority date: 1989-01-26
Filing date: 1990-01-26
Publication date: 1990-08-09
Also published as: AU5022690A; GB8901675D0; EP0455690A1; CA2008697A1; AU621201B2; ZA90606B

Abstract

A variety of genes may be used to express protein which inhibits full expression of selected characteristics of plants. These inhibit functions which are critical to full expression of a genetic characteristic. Also known as 'killer' genes, a particular area of interest is in the expression of a protein inhibitor of mitochondrial function leading to cell death and failure to produce viable pollen, thus imparting male sterility. When inserted as a module in a gene cascade which permits external control of expression, male fertility may be restored.

Description

Inhibition of Plant Cell Respiration

The present invention relates to a method of inhibiting respiration of a plant cell by use of a gene, which is expressible in plants, to inhibit mitochondrial function, hence disrupting full expression of a selected plant characteristic.

According to the present invention there is provided a method of inhibiting gene expression in a target plant tissue comprising stably transforming a plant cell of a type from which a whole plant may be regenerated with a gene construct carrying a tissue-specific or a development-specific promoter which operates'in the cells of the target plant tissue and a disrupter gene encoding a protein which is capable, when expressed, of inhibiting respiration in the cells of the said target tissue resulting in death of the cells. Preferably the disrupter gene is selected from:

(a) The mammalian uncoupling protein (UCP) cloned from mammalian (usually rat) brown adipose tissue.

(b) A mutated form of the gene for the β-subunit of F.-ATPase which has sequences added or deleted such that these changes result in the retention of the ability to assemble with other subunits but interfere with function as an ATP synthase... The ability of these altered subunits to assemble correctly will be important as the required phenotypic effect of their expression will depend on their competition with wild-type subunits for binding sites in the enzyme complex. Thus complexes containing non-functional subunits will only be weakly active and mitochondria harbouring these complexes will be non-functional. (c) A mutated, synthetic form of the oli 1 gene encoding subunit 9 of the F -ATPase. Mutations created as described at (b) above.

(d) A mutated form of a mitochondrial transit pre-sequence which malfunctions during transfer resulting, probably by blocking of receptor sites, in the disruption of protein transport to mitochondria.

(e) Gene constructs involving a fusion between the β-subunit gene from yeast and the β-glactosidase gene from E. coli,^" resulting in expression of a disrupting fusion protein.

Preferably the promoter is a tapetum-specific promoter or a pollen-specific promoter, so that on expression of the said disrupter protein therein the regenerated.plant is in male sterile. More preferably the said tapetum-specific promoter has the sequence shown in Figure 1 or 2 or 3 of the accompanying drawings.

Plasmids containing the DNA sequences shown in Figures 1, 2 and 3 have been deposited under the terms of the Budapest Treaty, details being as follows:

Plasmid pMSlO in an Escherichia coli strain RRl host, containing the gene sequence shown in Figure 1 herewith, and deposited with the National Collection of Industrial & Marine Bacteria on 9th January 1989 under the Accession Number NCIB 40090.

Plasmid pMS14 in an Escherichia coli strain DH5α host, containing the gene control sequence shown in Figure 2 herewith, and deposited with the National Collection of Industrial & Marine Bacteria on 9th

January 1989 under the Accession Number NCIB 40099.

Plasmid pMSlδ in an Escherichia coli strain RRl host, containing the gene control sequence shown in Figure 3 herewith, and deposited with the National Collection of Industrial & Marine Bacteria on 9th January 1989 under the Accession Number NCIB 40100.

The isolation and characterisation of these gene control sequences of this invention are described in full in a copending patent application.

The present invention also provides a plant having stably incorporated in its genome by transformtion a gene construct carrying a gene construct carrying a tissue-specific or a development-specific promoter which operates in the cells of the target plant tissue and a disrupter gene encoding a protein which is capable, when expressed, of inhibiting respiration in the cells of the said target tissue resulting in death of the cells.

The invention also provides a plant, particularly a monocotyledonous plant, and more particularly a corn plant, having stably incorporated within its genome a gene construct carrying a tissue-specific promoter which operates in the cells of the said target tissue and a disrupter gene encoding a protein which is capable of inhibiting respiration in the said cells of the said target tissue resulting in death of the cells.

These gene constructs may be used as a means of inhibiting cell growth in a range of organisms from simple unicells to complex multicellular organisms such as plants and animals. By the use of tissue- or cell-specific promoters, particular cells or tissue may be targeted and destroyed within complex organisms. One particular application intended for this invention is in the destruction of cells essential for male flower development, leading to male sterility.

The invention therefore provides a method of preventing or inhibiting growth and development of plant cells based on gene constructs which inhibit respiratory function. The technique has wide application in a number of crops where inhibition of particular cells or tissue is required.

Of particular interest is the inhibition of male fertility in maize for the production of Fl hybrids in situ. The concept of inhibition of mitochondrial function as a mechanism for male sterility arises from some previous research on T-type cytoplas ic male sterility in maize (cms-T) which has shown an association between the male sterile phenotype and mitochondrial dysfuction. Although a direct causal relationship has yet to be established between mitochondrial dysfunction and cms-T, an increasing body of evidence suggests that fully functional mitochondria, particularly in the tapetal cells, are essential. This is particularly critical during icrosporogenesis since the metabolic demands placed on the tapetal cells results in a 40-fold increase in mitochondrial number.

Thus we provide a number of negative mutations which act,upon mitochondria to uncouple oxidative phosphorylation. When specifically expressed in maize anther tissue these mutations will result in a male sterile phenotype.

The proposed disrupter protein, UCP, is instrumental in the thermogenesis of mammalian brown adipose tissue and exists as a dimer in the mitochondrial inner membrane forming a proton channel and thus uncoupling oxidative phosphorylation by dissipation of the proton electrochemical potential differences across the membrane. An alternative is the use of chimeric gene constructs in which domains are swapped, creating non-functional proteins. The target proteins here are the (3-subunit of F.-ATPase and subunit 9 of the F_Q- ATPase. During assembly of functional ATPase complexes, the altered chimeric subunits will complete for binding sites normally occupied by the naturally occurring subunits, particularly when the chimeras are over expressed compared with the endogenous genes. Mitochondrial function will be disrupted since F. and F_Q ATPase's assembled with altered subunits are likely to be weakly active or non-functional.

The method employed for transformation of the plant cells is not especially germane to this invention and any method suitable for the target plant may be employed. Transgenic plants are obtained by regeneration from the transformed cells. Numerous transformation procedures are known from the literature such as agroinfection using Agrobacterium tumefaciens or its Ti plasmid, electroporation, microinjection of plant cells and protoplasts, microprojectile transformation and pollen tube transformation, to mention but a few. Reference may be made to the literature for full details of the known methods. The development and testing of these gene constructs as disrupters of mitochondrial function in the unicellular organism, yeast, will now be described. A mechanism by which these gene constructs may be used to inhibit plant cell growth and differentiation in transformed plants will also be described. The object of these procedures is to use yeast as a model system for the identification and optimisation of gene constructs for expressing proteins which disrupt mitochondrial function. Plant cells will then be transformed with the selected constructs and whole plants regenerated therefrom.

The accompanying drawings are as follows: Figure 1 shows the DNA sequence of a tapetum-specific promoter, carried by plasmid pMSlO; Figure 2 shows the DNA sequence of a tapetum-specific promoter, carried by plasmid pMS14;

Figure 3 shows the DNA sequence of a tapetum-specific promoter, carried by plasmid pMSlδ; Figure 4 is a map of plasmid pCGSHO-UCP;

Figure 5 shows the mRNA sequence of mammalian uncoupling protein gene from plasmid pCGSHO-UCP (shown in Figure

4);

Figure 6 is a flowchart representation of the generation of a leu2, gall yeast strain;

Figure 7 is a table showing the effect of addition of galactose on the growth of BET9 and BET27 transformants;

Figure 8 shows results of growth curve analysis of BET9

(Figure 8A) and BET27 (Figure 8B) transformants grown on gly/cas medium over a period of 65 hours in the presence or absence of galactose;

Figure 9 is the growth curve analysis of rat UCP in strain BET9 grown on gly/cas medium over a period of 50 hours in the presence or absence of galactose; Figure 10 is the growth curve analysis of rat UCP in strain BET9 grown on raffinose medium over a period of

45 hours in the presence or absence of galactose;

Figure 11 illustrates the construction of plasmid

YIP/UCP from pKV49-UCP and YIp5; Figure 12 is a map of plasmid pGR208 (Figure 12A) and the sequence of oligonucleotides used to mutate the

(3-subunit gene of F.-ATPase (Figure 12B);

Figure 13 shows maps of plasmids pKV49 (Figure 13A) and

PKV49-UCP (Figure 13B); Figure 14 shows schematically the construction of a

(3-subunit/β-galactosidase fusion protein;

Figure 15 is a plasmid map of pKV49/BLZ;

Figure 16 is a plasmid map of pMSl0-5; and,

Figure 17 is a plasmid map of pBin/MSlO-UCP. The invention will now be illustrated by the following

Examples.

EXAMPLE 1

It was known from reports in the literature that the rat UCP gene inserted in the yeast/E. coli shuttle vector gave only low levels of expression of UCP. The yeast was Saccharomyces cerevisiae strain YM147 and the

UCP gene was available on plasmid pCGSHO-UCP.

Given the lack of useful expression levels with the wild type gene, modification of the rat UCP gene using site directed mutagenesis was carried out. the following modifications were made:

1. Introduction of a BamHI site seven nucleotides 5' to the AUG methionine initiation codon; 2. Modification of the sequence around the AUG methionine initiation codon to conform to the yeast consensus sequence ATAATG;

3. Deletion of an internal BamHI site; and,

4. Introduction of a BamHI site one nucleotide 3' to the TAG termination codon.

These modifications result in the deletion of the untranslated 5' and 3' rat UCP sequences as well as the introduction of a yeast consensus sequence at the methionine initiation codon. The 1.9 kb EcoRI/Pstl fragment from the plasmid pCGSHO-UCP (a map of the plasmid is shown in Figure 4 and the mRNA sequence of the UCP gene is shown in Figure 5) carrying the GAL10 promoter region and the rat UCP cDNA was cloned into the EcoRI/Pstl sites of Ml3mpl9 DNA. Sequencing of the resultant construct was carried out to ensure the correct structure.

Site directed mutagenesis was carried out according to the directions given in the Amersham (Trade Mark) mutagenesis kit using three different oligonucleotides as follows:

UCP-1 wild type CTCTGCCCTCCGAGCCAAGATGGTGAGTT mutant CTCTGCCCTCGGATCC(ATAATG)GTGAGTT

UCP-2 wild type TGCGACTCGGATCCTGGAACG mutant TGCGACTCGGTTCCTGGAACG

UCP-3 wild type ACCACATAGGCGACTTGGAG mutant ACCACATAGGATCCGACTTGGAG

Oligonucleotide UCP-1 was used to introduce the yeast consensus sequence (bracketed) which occurs around the methionine initiation codon, as well as the introduction of the BamHI cleavage site (underlined) .

Oligonucleotide UCP-2 was used to delete an internal BamHI site (underlined) .

Oligonucleotide "UCP-3 was used to introduce a BamHI site immediately after the TAG stop codon (underlined). These three mutations allowed the isolation of the entire UCP coding sequence on a 0.93 kb BamHI fragment. After selection of mutant clones the modified DNA was digested with BamHI. Three clones from twenty selected gave inserts of 0.93 kb upon digestion with BamHI.

Sequencing of the clones UCPSl and UCPS4 revealed that the UCP gene had been correctly modified with no unwanted changes present. The UCP gene was then transferred to the yeast expression plasmid pKV49 which allows expression of foreign genes in S. cerevisiae under the control of the strong PGK promoter and the GALl-10 UAS allowing induction/repression of the foreign gene according to whether or not galactose is present in the growth medium. The 0.93 kb BamHI fragment containing the modified UCP gene was cloned into pKV49 at the Bglll restriction site, resulting in the construct pKV49-UCP. TRANSFORMATION OF YEAST WITH pKV49-UCP CONSTRUCT a) Development Of Suitable Yeast Strain

For a recipient for the pKV49-UCP construct we needed a yeast strain carrying the appropriate markers for transformation and allowing induction of gene expression from the GALl-10 UAS while being unable to utilise glactose as a carbon and energy source (GALl, GAL2) . Such strains were generated by mating yeast strains YM147 and SF747. After selection of diploids on minimal plates containing uracil, the colonies were transferred to sporulating media. The resulting spores were grown on YDP plates prior to the resulting yeast colonies being characterised (Figure 6). Two new yeast strains BET9 (ura3, trpll, leu2, his3, gall) and BE27 (ura3, trpl, leu2, gall) were isolated, both of which are suitable for transformation with PkV49 based constructs. b) Yeast Transformation

Yeast strains BET9 and BE27 were transformed with pKV9 and pKV49-UCP DNA; transformants were selected using the appropriate auxotrophic selection (leu) and checked by plasmid isolation followed by restriction mapping. Single colonies from each of the four different transformants BET9/pKV49, BET9/pKV49-UCP, BE27/pKV49 and BE27/pKV49-UCP) were resuspended in sterile water prior to being spotted onto plating media containing a variety of carbon sources (Figure 7) both in the presence and absence of galactose. Results from these plate tests (Figure 7) indicated that on a few of the carbon sources used, the presence of both galactose and the pKV49-UCP construct resulted in poorer growth of the resulting yeast colonies. The greater effect on retardation of growth was observed with the glycerol/casamino (gly/cas) medium containing glactose for both pKV49-UCP transformants. Transformants either lacking the UCP gene or induced by galactose grew at the same rate as the untransformed BET9 and BE27 strains. GROWTH CURVE ANALYSIS As plating tests had indicated poor growth of pKV49-UCP transformants on gly/cas medium in the presence of galactose, growth curve analysis in liquid culture was carried out to determine more accurately the magnitude of the growth defect. The results in Figure 8 substitute the results of plating tests and indicate that neither the presence of pKV49-UCP DNA or galactose alone is sufficient to have any effect on the yeast cell growth rates, while the presence of both severely retards growth. As our initial results using the yeast strain YM147 transformed with the construct pCGSHO-UCP had not shown any significant growth defect on any of the tested carbon sources in the presence of galactose, it would appear that the modification o,f the UCP gene and/or the use of a different vector (pKV49) have resulted in an observable growth defect. ANALYSIS OF UCP EXPRESSION

As the growth curve analysis had indicated no detectable differences between the BE27 and BET9 transformants (Figure 8), it was decided only to use the BET9 transformants in subsequent experiments. Repeat growth analysis on gly/cas medium both in the presence and absence of galactose was carried out with the BET9 transformants. Cultures were allowed to grow for 47 hours to ensure that the same growth curve characteristics observed previously (Figure 9) were repeated. Cells were then harvested, total cell proteins were isolated and fractionated (in duplicate) by SDS-PAGE on a 10% polyacrylamide gel. One set of fractionated proteins were stained with Coomassie Blue to ensure equal loading of the proteins while the other set were transferred to nylon membrane and subjected to Western blot analysis using the rat UCP antibody. The Western blot showed two main features: 1) The comparative level of UCP expression between the BET9/pKV49-UCP transformant and the

VYl47/pCGSllO-UCP transformant reveals that the UCP expression has increased approximately 50-100 fold as a consequence of our modifications. 2) The yeast transformant which exhibits defective growth when grown on gly/cas medium in the presence of galactose also expresses substantial amounts of UCP.

It can be concluded from these results that the modification of the UCP gene and/or its subsequent cloning into the pKV49 vector has resulted in the increased level of UCP expression relative to the levels initially detected with the pCGSHO-UCP construct. Growth curve analysis indicates that the expression of UCP has an effect on the growth rates of yeast cells grown under certain conditions. As yet we have not been able to identify the specific effect that the increased levels of UCP expression have on yeast cell growth rates but preliminary results implicate a mitochondrial defect.

Growth curve analysis carried out in the raffinose medium (a fermentable carbon source which should not affect Gal regulation) of the BET9 transformants grown in the presence or absence of galactose indicate that the presence of both the UCP gene and galactose has no effect on growth rates (Figure 10). Western blot analysis of the proteins isolated from cells harvested during these growth curves reveals levels of UCP expression similar to those found in cells grown in gly/cas medium in the presence of galactose.

The UCP detected in BET9/pKV49-UCP transformants grown without added galactose is probably due to galactose residues released into the medium by hydrolysis of raffinose, possibly during the autoclaving. These observations indicate that the presence of UCP in yeast cells grown on a fermentable carbon source (no requirement for oxidative phosphorylation) has no effect on cell growth rates, while cells growing on the gly/cas medium (a non-fermentable carbon source) expressing UCP exhibit defective growth.

LOCATION OF UCP IN YEAST CELLS

Rat UCP is a major component of the mitochondrial inner membrane of brown adipose tissue. Unlike many other polypeptides found in the inner membrane it does not contain a cleavable signal sequence, the targeting information being encoded internally within the amino acid sequence of the protein. As our results indicate that the expressed UCP has an effect on the rate of yeast cell growth then it is important to determine the precise location of the protein expressed in yeast cells. Initial Western blot analysis of total mitochondrial proteins shows the UCP expressed by the pKV49-UCP transformant to be located in the mitochondrial fraction.

Subsequent mitochondrial fractionation revealed that the majority of the UCP is located in the inner membrane fraction of yeast mitochondria. Although some of the UCP appears to be located in the inner-membrane space, this observation is most likely due to contamination of this fraction with some of the inner membrane fraction. Similar results have been obtained with the location of the β-subunit of the F.-ATPase complex from yeast cell mitochondria. The β-subunit which is a component of the inner membrane is also detected in our inter-membrane space preparations. However, these results do show that the targeting information within the rat UCP is sufficient to target the UCP to the inner membrane of yeast mitochondria where it could function a an uncoupler protein. UCP TRANSCRIPT ANALYSIS

RNA has been isolated from many of the growth curve experiments described previously. We are currently carrying out Northern blot analysis in order to determine whether the patterns of UCP expression are reflected by the UCP transcript signals EFFECT OF COPY NUMBER ON UCP EXPRESSION The transformation of yeast cells with shuttle vectors containing the origin of replication from the yeast 2 μm circle, such as pKV49-UCP, results in these plasmids being present at approximately 40-50 copies per cell. Consequently any foreign gene carried by the plasmid will be present at the same relatively high copy number which may result in the expression of the foreign protein at a higher level than would be seen- for a gene present at a low copy number. We have therefore attempted to lower the copy number of the UCP gene by integrating it into the yeast chromosome at a single site resulting in a genetically stable, single-copy transformant. The vector YIp5 (Figure 11) is an integrating yeast vector carrying the ura3 gene; it is unable to replicate autonomously in yeast. The 1.8 kb EcoRI/Sall fragment from pKV49-UCP containing the rat UCP gene along with the PGK promoter and GAL UAS was cloned into the EcoRI/SALI sites of YIp5 DNA. The resultant plasmid UIP-UCP (Figure 11) was checked by restriction enzyme mapping to ensure the UCP gene was correctly inserted . The YIP-UCP plasmid was cut with the restriction enzyme EcoRV (which cuts in the middle of the URA3 gene (Figure 11) and the linearised YIP/UCP DNA was used to transform the yeast cell lines BET9 and BE27. Transformants were initially screened on minimal plates by selecting for uracil prototrophy and after 7-10 days two transformants from each cell line were streaked out onto YPD plates (non-selective). The transformants were then subjected to four consecutive periods of growth on non-selective medium. One hundred colonies from each of the original four transformants were then replica plated onto both non-selective (YDP) and selective media (minimal plants + ura). All colonies grew on the ^"selective media indicating that the URA3 gene, which is genetically linked to the UCP gene (Figure 11), had been integrated to the yeast cell chromosome. Chromosomal DNA was isolated from each of the four transformants, digested with the restriction enzyme EcoRI and fractionated on a 0.08% agarose gel. Southern blot analysis using a labelled UCP probe indicated that the UCP gene is present in the yeast^" chromosome of all four transformants. Western blot analysis using the rat UCP antibody will show the level of UCP expression in these transformants. Growth curve analysis of these transformants grown in the presence of galactose shows that they may have growth inhibition consistent with a mitochondrial defect. EXAMPLE 2 MODIFICATION OF THE β-SUBUNIT OF F-^ATPASE The second approach we have ¥aken to introducing mutations affecting mitochondrial function is the directed modification of functional mitochondrial proteins which when expressed in yeast might be expected to interfere with the generation of ATP. The protein chosen for this approach is the β-subunit of the F--ATPase complex. The DNA sequence of the yeast β-subunit gene is known and the gene has been independently cloned and sequenced in our laboratory 918).

The F.ATPase portion of ATP synthase catalyses the terminal step of oxidative phosphorylation F- is an assembly of five different polypeptides designated α, β, γ, δ and ε. Experiments carried out by Parsonage et al on modification of the β-subunit of F.-ATPase from E. coli identified specific amino acid residues of the β-subunit that appear to be very important for catalysis of both ATP synthesis and hydrolysis. Two mutations in particular were shown to result in greatly impaired catalysis without causing major structural perturbation of the F..-ATPase. One of these mutations resulted from changing the strongly conserved lysine residue occurring in the catalytic nucleotide-binding domain at position 155 to a gluta ine residue while the other mutation resulted from changing the methionine residue at position 209 to a leucine residue. Both of these mutations have been reposed to exert their effect by the prevention of confirmational changes required from the catalytic cooperativity in the F. complex. As the assembly of these mutated β-subunit proteins into the F.-ATPase is not affected, then it was felt that similar mutants of the β-subunit in yeast might compete for assembly into F.-ATPase. It was thought that the result of having both wild-type and mutated β-subunits in the same F.-ATPase would perhaps result in impaired catalysis resulting in a decrease in ATP production and retarded cell growth.

The β-subunit of F.-ATPase from a wide variety of sources has been shown to by highly conserved at the amino acid sequence and comparison of the S cerevisiae β-subunit amino acid sequence with that from E. Coli confirms that the lysine and methionine residues shown by Parsonage et al to be very important for catalytic activity are conserved, with the lysine and methionine residues occurring at positions 196 and 255 respectively on the yeast β-subunit sequence.

In order to carry out SDM the wild-type β-subunit gene from yeast was isolated from the plasmid pGR208 (Figure 12) as an EcoRI/BamHI fragment which was cloned into Ml3mpl9. Two mutated β-subunit genes were constructed: mutant BBl has both the Met255 and Lysl96 converted to isoleucine and glutamine respectively while mutant BB2 has only the lysine to glutamine mutation (Figure 12). Following sequence analysis to ensure correct mutagenesis with no unwanted mutations, the mutated β-subunit genes were removed from mp!9 by EcoRI/BamHi digests. The fragments containing the genes were then'blunt-ended and ligated to Bglll digested pKV49 (Figure 13) which had previously been blunt-ended. We have both mutated β-subunit genes cloned into pKV49 (pKV49-BBl and pKV49-BB2) and have transformed the yeast strain BET9 with both these constructs. Growth curve and plate growth from both mutated β-subunit transformants show tht the transformants have altered growth characteristics which are consistent with a mitochondrial defect.

Concurrently with the transformation of strain BET9 with the mutated β-subunit genes, gene disruption may be used to construct a derivative of strain BET9 which will fail to synthesize β-subunit. The resultant strain will therefore be unable to grow on non-fermentable carbon sources although it will be easily maintainable on a fermentable carbon source such as glucose. Transformation of this strain with plasmids bearing the mutated β-subunit genes, followed by measuring the transformants' growth characteristics on a non-fermentable carbon source, shows that the altered β-subunit is unable to support oxidative phosphorylation. EXAMPLE 3 FUSION PROTEINS

An alternative strategy for selectively perturbing mitochondrial function is the expression of a fusion protein which results in either poor or no yeast cell growth. The candidate fusion protein chosen from this project contains the N-terminal region of the yeast ATP synthase β-subunit fused to most of β-galactosidase from E. coli and has been constructed by gene fusion (Figure 14). This β-subunit/β-galactosidase fusion protein has already been shown to be targeted to the inner membrane of yeast mitochondria 921) and cells expressing this fusion protein appear to be unable to grow on a non-fermentable carbon source. In the presence of the fusion protein the transducing capacity of the mitochondrial membrane as measured by the 32P-ATP exchange reaction is only 9% of that measured in the absence of the fusion protein. As yet the mechanism of this description has not been evaluated but the gene fusion is thought to produce a protein which becomes trapped in the inner membrane and interferes with function(s) essential for respiratory growth. CONSTRUCTION OF THE ATP2/LacZ GENE FUSION The plasmid pGR208, which contains the yeast ATP2 DNA encoding ATP synthase β-subunit gene (Figure 12), was digested with EcoRl plus BamHi resulting in the release of a l.lkb fragment coding for the first 350 amino acids of the β-subunit protein. pMURl720 is a pUC8 based plasmid which contains a Lacz gene contained within an EcoRI/Narl fragment (Figure 14). Cloning of the l.lkb EcoRI/BamHi DNA fragment coding for the first 350 amino acids of the yeast β-subunit protein into the EcoRI/BamHi sites of pMUR1720 (Figure 14a) results in an in-frame fusion between the 350 amino acids of the β-subunit and the entire (minus the first eight amino acids) LaZ protein (Figure 14). The entire β-subunit/LacZ gene fusion is now contained on the 4.3kb EcoRI/Narl fragment in construct pMURl720-BLZ (Figure

14). This 4.3kb EcoRI/Narl fragment is currently being cloned into the pKV49 vector resulting in the pKV49-BLZ construct (Figure 15) which can be used to transform the yeast strains BET9 and BE27 and show that when induced by galactose growth defects consistent with mitochondrial inhibition arise. EXAMPLE 4

Construction of a promoter fusion between the MS10 gene and the UCP gene The 1830 bp Hindlll to BamHI fragment from pMSlO was ligated into the binary plant transformation vector Binl9 previously cut with Hindlll and BamHI.

Following ligation the resultant plasmid was cut with BamHI and ligated to the 930 bp UCP BamHI fragment from plasmid pUC/UCP (s derivative of pUCl9 contyaininbg the modified UCP gene cloned at the BamHI site) to construct a fusion between the MS10 gene promoter and the UCP gene. Finally the nos 3' terminator obtained as a 250 bp Sstl- EcoRl fragment from vector pTAKl wasligated into the MS10-UCP construct previouslt cut with Sstl and EcoRl.

The resulting plasmid is termed pBin/MSlO-UCP and contains the MS10 promoter, the UCP gene, nos 3^r terminator expression cassette located between the right and left border sequences of Agrobacterium T-DNA allowing efficient transformation into tobacco cells. EXAMPLE 5 Transformation of tobacco plants with pBin/MSlO promoter gene constructs

The recombinant vector pBin/MSlO-UCP was mobilised from E Coli (TG-2) onto Agrobacterium tumefaciens (LBA4404) in a triparental mating on L-plates with E_ Coli (HB101) harbouring pRK2013. Transconjugants were selected on minimal medium containing kanamycin (50 g/cm ) and streptomycin (500,vg/cm ).

L-Broth (5 cm ) containing kanamycin at 50 g/cm was inoculated with a single Agrobacterium colony. The culture was grown overnight at 30°c with shaking at 150 rpm. This culture (500/1) was inoculated into L-Broth containing kanamycin (50/ g/cm ) and grown as before. Immediately before use the Agrobacteria were pelleted by spinning at 3000 rpm for 5 minutes and suspended in an equal volume of liquid Murashige and Skoog (MS) medium. Feeder plates were prepared in 9 cm diameter petri dishes as follows. Solid MS medium supplemented with 6-Benzyl-aminopurine (6-BAP) (1 mg/1) and 1-Naphthaleneacetic acid (NAA) (0.1 mg/1) was overlaid with Nicotiana tabacum var Samsun suspension culture (1 cm ) . One 9 cm and one 7cm filter paper discs were placed on the surface.

Whole leaves from tissue culture grown plants were placed in the feeder plates. The plates were sealed with "Nescofilm" and incubated overnight in a plant growth room (26°C under bright fluorescent light). Leaves from the feeder plates were placed in

Agrobacteria suspension in 12 cm diameter petri dishes

2 and cut into 1- 1.5 cm sections. After 20 minutes tht leaf pieces were returned to the feeder plates which were sealed and replaced in the growth room. After 48 hours incubation in the growth room the plant material was transferred to MS medium supplemented with 6-BAP (1 mg/1), NAA (0.1 mg/1), carbenicillin (500//g/cm ) and kanamycin (100 //g/cm ), in petri dishes, the petri dishes were sealed and returned to the growth room.

Beginning three weeks after inoculation with Agrobacterium, shoots were removed from the explants and placed on MS medium supplemented with carbenicillin (200 _//g/cm 3) and kanamycin (100 g/cm3) for rooting.

Transformed plants rooted 1-2 weeks after transfer.

Following rooting, transformed plants were transferred to pots containing soil and grown in the glasshouse. Roughly^"one month after transfer the plants flowered.

The anthers of the tobacco plants containing the pBin/MSlO-UCP construct were were assayed for expression of the UCP gene by Northern blotting of RNA samples, and the effect of UCP expression on pollen development determined.

Claims

A method of inhibiting gene expression in a target plant tissue comprising stably transforming a plant cell of a type from which a whole plant may be regenerated with a gene construct carrying a tissue-specific or a development-specific promoter which operates in the cells of the target plant tissue and a disrupter gene encoding a protein which is capable, when expressed, of inhibiting respiration in the cells of the said target tissue resulting in death of the cells.

2. A method according to claim 1 in which, the disrupter gene is the mammalian uncoupling protein (UCP) gene.

A method according to claim 1 in which, the disrupter gene is a mutated form of the gene for the β-subunit of F--ATPase which has sequences added or deleted such that these changes result in the retention of the ability to assemble with other subunits but interfere with function as an ATP synthase.

4. A method according to claim 1 in which, the disrupter gene is a mutated, synthetic form of the oli 1 gene encoding subunit 9 of the F -ATPase.

5. A method according to claim 1 in which, the disrupter gene is a mutated form of a mitochondrial transit pre-sequence which malfunctions during transfer resulting in the disruption of protein transport to mitochondria.

A method according to claim 1 in which, the disrupter gene is a gene construct carrying a fusion between the β-subunit gene from yeast and the β-glactosidase gene from E. coli, resulting in expression of a disrupting fusion protein.

A method as claimed in claim 1, in which the promoter is a tapetum-specific promoter or a pollen-specific promoter, so that on expression of the said disrupter protein therein the regenerated plant is in male sterile.

A method as claimed in claim 2, in which the tapetum-specific promoter has the sequence shown in Figure 1 or Figure 2 or Figure 3 of the accompanying drawings.

9. The plasmid designated pBin/MSlO-UCP having the structure shown in Figure 17 of the accompanying drawings.

10. A plant transformation vector comprising

Agrobacterium tumefaciens harbouring the plasmid pBin/MSlO-UCP claimed in claim 9.

11. A plant having stably incorporated in its genome by transformation a gene construct carrying a gene construct carrying a tissue-specific or a development-specific promoter which operates in the cells of the target plant tissue and a disrupter gene encoding a protein which is capable, when expressed, of inhibiting respiration in the cells of the said target tissue resulting in death of the cells.