WO2003054201A1 - Method for the transformation of vegetable plastids - Google Patents

Abstract

The invention relates to new methods for producing transgenic plants with genetically modified plastids and the transgenic plants produced by said methods.

Description

Process for the transformation of plant plastids

description

The present invention relates to new processes for the production of transgenic plants with genetically modified plastids, and to the transgenic plants produced by these processes.

The aim of biotechnological work on plants is the production of plants with advantageous new properties, for example to increase agricultural productivity, to increase the quality of food or to produce certain chemicals or pharmaceuticals.

Piastids are organelles within plant cells with their own genome. They have an essential function in photosynthesis as well as amino acid and lipid biosynthesis. The plastid genome consists of a double-stranded, circular DNA with an average size of 120 to 160 kb and lies - e.g. in leaf cells - with approximately 1900 to 50000 copies per cell before (Palmer (1985) Ann Rev Genet 19: 325-54). A single plastid has a copy number of approx. 50 to 100. The term plastids encompasses chloroplasts, proplastids, etioplasts, chromoplasts, amyloplasts, leukoplasts and elaloplasts (Heifetz P (2000) Biochimie 82: 655-666). The different forms can be converted into one another and all arise from the proplastids. Therefore, all forms of the plastids contain the same genetic information. In the literature, however, green cells which contain chloroplasts as a form of expression are preferably used as the starting material for the transformation of plastids.

In addition to tobacco, the plastid transformation in potatoes (Sidorov VA et al. (1999) Plant J 19: 209-216; WO 00/28014),

Petunia (WO 00/28014), rice (Khan MS and Maliga P (1999) Nature

Biotech 17: 910-915; WO 00/07431; US 6,153,813), Arabidopsis

(Sikdar SR et al. (1998) Plant Cell Reports 18: 20-24;

WO 97/32977) and rapeseed (WO 00/39313) (review article: Bogorad L (2000) TIBTECH 18: 257-263). Recently, plastoplastome tomato plants have also been described (Ruf S et al. (2001) Nature

Biotech 19: 870-875).

It is of great economic interest for plant biotechnology to develop efficient methods for plastid transformation (McFadden G (2001) Plant Physiol. 125: 50-53). The stable transformation of plastids of higher plants is one of the major challenges.

The technique of non-targeted (illegitimate) DNA insertion that is often used when inserting into nuclear DNA has the disadvantage in plastid transformation that it is very likely that an essential gene will be struck on the gene-tight plastid genome, which would often be lethal for the plant , A targeted insertion of foreign DNA is therefore advantageous in plastids.

Various methods for targeted insertion into the plastid genome have been described. The plastid transformation in green algae was first described (Boynton JE et al. (1988) Science 240: 15 1534-1538; Blowers AD et al. (1989) Plant Cell 1: 123-132), later in higher plants such as tobacco (Svab Z et al. (1990) Proc Natl Acad Sei USA 87: 8526-8530).

EP-A 0 251 654, US 5,932,479, US 5,451,513, US 5,877,402,

20 WO 01/64024, WO 00/20611, WO 01/42441, WO 99/10513, WO 97/32977, WO 00/28014, WO 00/39313 describe methods and DNA constructs for transforming plastids of higher plants, the DNA to be transformed is incorporated into the piastome (piastid genome) via homologous recombination ("double crossover"). in the

25 general homologous regions of 1000 bp or more are used on each side of the sequence to be inserted. This very quickly leads to large vectors, whose handling is not very comfortable. In addition, the efficiency of the transformation decreases. The decreases with increasing length of the foreign DNA to be integrated

30 Efficiency of homologous recombination. Another disadvantage is the fact that a homologous area must be identified for each plant species that can be used for the process of DNA integration by means of double crossover. WO 99/10513 claims to identify an intergenic DNA sequence

35, which is sufficiently homologous between the genomes of the chloroplasts of many higher plants and can thus act as a universal target sequence. However, it has not been proven that this vector can be successfully used in species other than tobacco. On the contrary: In WO 01/64024 the same inventor 0 adapts the transformation vector to non-tobacco plant species by using homologous DNA sequences isolated therefrom. Since only a few recombination events result in all of the methods described above, a selection of the recombinant plastid DNA molecules is necessary. 5 The plastid DNA of higher plants is available in up to several thousand copies per cell. To achieve a stable integration of the foreign DNA, all copies of the plastid DNA must be changed equally. In the case of plastid transformation, one speaks of having reached the homotransplastomic state. This state is achieved through a so-called "segregation and sorting" process, in which the plants are kept under constant selection pressure. Due to the persistent selection pressure, those plastids in which many copies of the plastid DNA have already been changed are enriched during the division of the cells and plastids. The selection pressure is maintained until the homotransplastoma is reached (Guda C et al. (2000) Plant Cell Reports 19: 257-262). It is a major challenge to modify all copies of the plastid genome in order to obtain homotransplastome plants that have stably incorporated the foreign gene into their plastid genome without the addition of a selection agent for generations (Bogorad L (2000) TIBTECH 18: 257-263 ). In addition to the constant selection pressure, repeated regeneration of already transformed tissue ensures that the homotransplastomic state is reached (Svab Z and Maliga P (1993) Proc Natl Acad Sei USA 90: 913-917). However, this limits the plant material that is available for plastid transformation. It is therefore proposed to couple the transgene with another gene that is essential for the survival of the plant. In general, tobacco is regarded as the only plant species whose plastids can be transformed efficiently. The limitation of the technique is seen above all in the currently available tissue culture and the regeneration protocols (Ruf, S. et al. 2001 Nature Biotech. 19: 870-875). Tissue culture techniques and selection processes are usually not universally applicable to all plant species and represent a significant limitation of plastid transformation, which particularly affects the transferability of the method to species other than tobacco (Kota M et al. (1999) Proc Natl Acad Sei USA 96: 1840-1845). A recently published transformation of tomato plastids is based on modifications in the regeneration and selection scheme (Ruf S et al. (2001) Nature Biotech 19: 870-875), which, however, are costly and time-consuming. Another approach aims to reduce the number of plastids per cell and the number of DNA molecules per plastid so that fewer DNA molecules have to be modified (Bogorad L (2000) TIBTECH 18: 257-263). Overall, the selection and segregation processes are very time consuming.

WO 99/10513 describes a method in which a plastid ORI (origin of replication, origin of replication) is located on the plasmid to be transformed, so as to determine the number of to increase the integration into the plastid genome available copies of the vector to be transformed (Guda C et al. (2000) Plant Cell Reports 19: 257-262).

The need to improve the technique of plastid transformation is also mentioned in Heifetz and Tuttle (Heifetz P and Tuttle AM (2001) Curr Opin Plant Biol 4: 157-161). WO 00/32799 teaches to increase the efficiency of plastid transformation by using plants with enlarged plastids. This results in a large surface area of the plastids, through which the DNA to be transformed can penetrate the plastids more easily. The mechanism of DNA integration also takes place here by means of conventional homologous recombination as in the methods described above.

Homologous recombination itself is not discussed as a starting point for further optimization options, since it is considered to be high in plastids and therefore not limiting (Bock R and Hagemann R (2000) Progress in Botany. 61: 76-90; Maliga P et al. ( 1994) Homologous recombination and integration of foreign DNA in plastids of higher plants. In: Homologous recombination and gene silencing in plants. Paszkowski J ed., Kluwer Acade ic publishers, p.83-93; Carrer H and Maliga P (1995) Bio / Technology 13: 791-794).

In general, optimized tissue and regeneration techniques are favored as an approach, but they have the disadvantage that they are not universally applicable to all plant species, but must be adapted to their characteristics. Only two works suggest, also by improving the

Frequency of insertion of the DNA to be transformed into the DNA of organelles to improve the transformation of these. In one case, no technical teaching is described on how to improve the frequency of integration into plastid DNA (van Bei AJE et al. (2001) Curr Opin Biotechnol. 12: 144-149). Odom OW et al. (2001 Mol. Cell Biol. 21: 3472-3481) generally suggest increasing transformation efficiency by using rarely cutting restriction endonuclease.

Depending on the location and orientation of the corresponding recognition sequences, sequence-specific recombinases cause an insertion, excision, translocation or inversion of nucleic acid fragments. There is a difference between the families of integrases, such as Cre or Flp, and the family of resolvases / invertases, such as that ΦC31 recombinase, the TP901-1 recombinase, the gin recombinase, the Hin recombinase etc.

These enzymes usually recognize relatively short, specific nucleic acid sequences that are used both for recognition and for recombination. Examples include the Cre recombinase (Sternberg and Hamilton (1981) J Mol Biol 150: 467-486), the Flp recombinase (Brosch et al. (1982) Cell 29: 227-234) and the R recombinase (Matsuzaki et al. ( 1990) J Bacteriology 172: 610-618).

The Cre recombinase allows the bacteriophage PI to obtain a monomeric phage genome during its life cycle by cutting out the sequence duplications generated during the replication by the recombinase (Guo et al. (1997) Nature 389: 40-46). Cre is a 38 kD protein and catalyzes the recombination between two 34 base pair sequences (loxP).

loxP 5 '-ATAACTTCGTATA GCATACAT TATACGAAGTTAT-3'

LoxP consists of two 13 base pairs of palindromic sequences separated from an 8 base pair core sequence ("spacer"). The “repeat” sequences act as binding sequences of the Cre recombinase, whereas the sequence cut (“crossing of the sequences to be recombined”) takes place in the “spacer” region. A transition complex is presumably formed from four Cre molecules and two loxP sequences (Guo et al. (1997) Nature 389: 40-46). The asymmetry of the "spacer" region determines the direction of the recombination. If the two recombination sequences are aligned, integration or excision takes place. If the orientation is opposite, inversion occurs (Yang and Mizuuchi (1997) Structure 5: 1401-1406). Cre is active in a variety of organisms, including plants (Albert et al. (1995) Plant J 7: 649-659; Dale and OW (1990) Gene 91: 79-85; Odell et al. (1990) Mol Gen Genet 223: 369-378). Cre can effect a sequence-specific insertion, for example in yeast or mammalian cells (Sauer and Henderson (1990) New Biol 2: 441-449).

The Flp inversion system of the yeast S. cerevisiae (2μ plasmid) is based on the use of an Int-Flp recombinase. Flp recognizes a region - called FRT - consisting of two inverted "repeat" sequences, each 13 base pairs long, separated by a sequence (spacer) of 8 bp. A third repeat, which is still present in the natural sequence, does not appear to be essential for the functionality. An inverted localization of repeats # 1 and # 2 is essential. The offset endonucleolytic cut is made within the 8 bp sequence. The length of the spacer is important. However, an insertion of a further 10 bp can be tolerated. The system is based on two accompanying target sequences.

# 1 spacer # 2 # 3

FRT 5-GAAGTTCCTATaC TTTCTAGA GAATAGGAACTTC [C GAATAGGAACTTC] -3 '

Like Cre, Flp is also functional in various systems, including in the plant cell nucleus (Lyznik et al. (1993) Nucl Acids Res 21: 969-975).

Many recombination systems are of course reversible. When using recombinases such as Cre or FLP, the equilibrium of the reaction lies far on the side of the excision. Therefore, these systems - without further modification - are not suitable for the integration of foreign DNA, but only for excision (e.g. marker excision). The Cre-lox system of phage PI was used to remove the selection marker aadA from the plastid genome (Hajdukieicz PTJ et al. (2001) Plant J 27: 161-170; Corneille S et al. (2001) Plant J 27: 171-178;

WO01 / 21768; WO01 / 29241). So far, there are no descriptions of the benefits of recombinases in intermolecular reactions, in particular for the integration of foreign DNA, in plastids. Recombinant systems independent of the host organism have so far only been described in plastids in connection with a deletion of defined sequences from the piastome.

The use of sequence-specific recombinase systems for the integration of heterologous sequences into the chromosomal DNA (including from plants) has been described. In contrast to the excision, which is intramolecular, this is an intermolecular event. An attempt was made to increase the integration rate compared to the excision rate by creating at least one hybrid recombinase recognition site in the event of a recombination event between two recognition regions that are not identical but are still used by the recombinase and which can no longer or only be used very inefficiently by the corresponding recombinase , This suppresses the excision of the DNA fragment just integrated (Sauer B (1994) Curr Opin Biotechnol 5: 521-527) and stabilizes the integration (Albert H et al. (1995) Plant J 7: 649-659). Something similar is achieved by targeted, time-limited expression of the corresponding recombinase - such as Cre or Flp (Metzger D and Feil R (1999) Curr Opin Biotechnol 10: 470-476; Albert H et al. (1995) Plant J 7: 649-659). Current research on optimizing plastid transformation is not aimed at optimizing homologous recombination, but rather, for example, at improved selection markers, improved selection and regeneration techniques, etc.

The recombinase of the phage ΦC31 has already been used for biotechnological applications in the past. No. 5,190,871 describes a method by means of which a plasmid is inserted into the genome of a streptomycete with the aid of the recognition site-specific integration function of the phage PhC31. The integration takes place in the recognition region attB, which is naturally present in these host cells. WO 00/11155 uses recombinases to incorporate foreign DNA into eukaryotic genomes. The recombinases are expressed under the control of a eukaryotic promoter. The recombination takes place between one

Recognition region on the plasmid to be transformed and pseudo-recognition regions that are native to the genome of the eukaryote. In WO 00/11155, a recognition region attB or attP is introduced into the genome of a eukaryotic cell. This is only shown for animal cells and there is no technical disclosure that and how one could introduce such recognition regions in organelles. In contrast to the nuclear genome, which was considered in the addressed invention by having only one (heterozygous) to two (homozygous) genome copies, there are many plastids per cell and many copies of DNA per plastid.

While WO 00/11155 only carries out investigations in bacteria and animal cells, WO 01/07572 describes the use of such recombinases also in the nuclear genome of plants. Here, too, it is not shown that such recombinases recognize the DNA structure of the plastid DNA and that an efficient application is possible there due to the large number and spatial separation of these DNA molecules. In fact, the inventor was not faced with such a task at all, since he was looking for a method which was an alternative to the homologous recombination occurring in the nuclear genome with only very low frequencies.

Different recombinases prefer different conformations of the DNA that they use as substrates (Sadowski PD (1986) J Bacteriol 165: 341-347). It is therefore unclear whether and which recombinases - in addition to the Cre-lox system - work in plastids. So far, no recombinases of the resolvase / invertase family have been used in plastids of higher plants. Since the nucleosomes are constructed completely differently than the chromosomes in the cell nucleus, it was also not possible to use certain recombinases in the plant cell nucleus assume that these can also be used as a substrate with plastid DNA.

So-called site-specific integration has so far only been used in the core genome of plants and animals. The aim of the work described in the prior art was to provide an alternative method to homologous recombination for sequence-specific integration into the core genome. In the core genome of higher eukaryotes, systems such as Cre or Flp are at most as efficient as the illegitimate recombination that normally occurs there, but mostly much more inefficient (Sauer B (1994) Curr Opin Biotechnol 5: 521-527). In contrast to the nuclear genome, the piastome is available in numerous, usually several thousand copies per cell, so that particularly high requirements must be placed on the efficiency of the insertion.

In contrast to plastids, insertion into the core DNA by means of homologous recombination is problematic and usually takes place by means of random illegitimate integration. This shows that techniques established in the core genome cannot necessarily be transferred to the plastids. In contrast to the situation in the core, plastids of higher plants are almost exclusively and highly efficiently integrated via homologous recombination (Bock R and Hagemann R (2000) Progress in Botany 61: 76-90; Maliga P et al. (1994) Homologous recombination and Integration of foreign DNA in plastids of higher plants. In Homologous recombination and gene silencing in plants. Paszkowski J, ed. (Kluwer Academic publishers), pp. 83-93).

The efficiency of homologous recombination for DNA integration into the piastome was generally not regarded as a limiting factor and - on the contrary - was classified as uncritical, so that the use of recombinases to integrate foreign DNA into the plastid DNA appears to be superfluous.

As the processes and problems in plastid transformation described above clearly highlight, the provision of new processes for the production of homotransplastomeric plants is a long-standing unsolved need in plant biotechnology. Another need is to avoid antibiotic or herbicide selection markers for regulatory and consumer acceptance reasons. No method has yet been described for plastid transformation that can do without such a selection marker. The task therefore arose to develop new methods which ensure efficient integration of foreign DNA into all copies of the plastid DNA and thus enable efficient selection of appropriate homotransplastomeric plants. In addition, there is the task of providing efficient methods for the integration of large DNA fragments in particular into the piastome in a single transformation step. This object has been achieved by providing the integration method according to the invention.

A first subject of the invention relates to a method for the sequence-specific integration of a DNA sequence into the plastid DNA of a plant organism or cells derived therefrom, characterized in that

a) the plastid DNA molecules of said plant organism or cell derived therefrom contain at least one recombination sequence R1 and

b) i) at least one transformation construct comprising an insertion sequence flanked at least on one side by at least one further recombination sequence R2 and

ii) at least one sequence-specific recombinase suitable for inducing recombination events on said recombination sequences R1 and R2

in at least one plastid of said plant organism or cells derived therefrom, and

c) the insertion sequence is inserted into the plastidic DNA with recombination of R1 with R2, and

d) plants or cells are isolated in which the insertion sequence has been inserted into the plastid DNA molecules.

The system surprisingly enables a substantial one

Increased efficiency in the production of predominantly homo-transplastomer plants. This increases both the efficiency of insertion into the plastid DNA and the efficiency of the selection process of predominantly homotransplastomeric plants. It is particularly advantageous that when using the method according to the invention, homologous areas in the transformation vectors can be dispensed with. This will make the vectors smaller and easier to handle. Furthermore, the efficiency of plastid transformation and thus the overall efficiency of the method is increased. The latter is of particular interest if large DNA fragments are to be introduced into the piastome as part of the insertion sequence, which is the case, for example, if several genes are introduced. As a result, the method according to the invention is superior to the insertion described in the prior art by means of homologous recombination (eg double crossover), since in the latter the efficiency of the transformation (and thus the overall efficiency) is reduced by the additionally required homology areas. The present invention thus solves the problem of integrating large DNA fragments (> 4 kb) in a single transformation step with sufficiently high efficiency.

Advantages of the procedure described here are: A master plant - i.e. a plant which contains a recombination sequence R1 in your piastome and which can function as a starting plant for the method according to the invention - be created in a corresponding manner. Unmodified plants can also be used as master plants, provided that their piastome naturally contains a recognition sequence for a recombinase. The plastids of the aforementioned master plants can then be transformed using the host-independent integration mechanism according to the invention. This can increase the efficiency of plastid transformation in the plant species, in which the integration of the foreign DNA into the piastome is a limiting factor. Only a few base pairs long sequence sections are required on the transformation vectors, which determine the integration site in the piastom. In contrast, sequence cross-sections with extensive homology to the piastome of the host plant are required for integration by means of double crossover. The transformation vectors of the present invention can thus be kept relatively small. In addition, the host-independent integration is largely independent of the length of the DNA to be introduced into the master plant. Very many genes or very large genes can easily be integrated into the organelles of such a master plant using the method described here.

Compared to the transformation of the cell nucleus, the transformation of plastids has numerous advantages. These include:

a) While homologous recombination into nuclear DNA is difficult to achieve, plastid DNA can easily be integrated at a predefined location using double crossover, a form of homologous recombination. Positions effects or "genesilencing" that can be found in core transformations due to the illegitimate integration into a non-predefined locus are avoided.

b) Very high levels of expression can be achieved, presumably due to the high copy number of the plastid DNA.

c) The DNA of the plastids is usually only maternally inherited in higher plants, so that the introduced foreign DNA cannot be spread via pollen and cross-breeding can thus be effectively prevented.

d) The prokaryotic nature of the plastids enables the expression of genes in the context of a polycistronic operon structure. It is therefore not necessary to equip each gene to be expressed with its own promoter, etc. This makes it easier to insert many genes at once, for example to insert entire biosynthetic pathways into the plastids.

“Plastid” means proplastids and all organelles resulting therefrom, such as chloroplasts, etioplasts, chromoplasts, amyloplasts, leukoplasts, dermatoplasts and elaloplasts (Heifetz P (2000) Biochimie 82: 655-666).

"Piastom" means the genome, that is, the entirety of the genetic information, of a plastid.

"Homotransplastoma" means a transplastomic and homoplastic state.

"Transpiastome" means, for example with respect to a plant, cell, tissue, plastid or plastid-DNA, all such forms of the aforementioned that have been obtained by genetic engineering methods and comprise a plastid DNA that has been modified by genetic engineering methods, the modification exemplifying substitutions, May include additions, deletions, inversions or insertions of one or more nucleotide residues.

"Heteroplastoma" means the presence of a mixed population of different plastid DNAs within a single plastid or within a population of plastids within a plant cell or tissue.

"Homopiastoma" means a uniform population of plastid DNA within a single plastid or within a population of plastids within a plant cell or Tissue. Homoplastomas Cells, tissues or plants are genetically stable because they only contain one type of plastid DNA, ie they generally remain homoplastomas even if the selection pressure no longer persists. Offspring obtained by selfing are also homoplastic.

In the context of this invention, "predominantly homoplastom" or "predominantly homotransplastom" means all those plants or cells in which the proportion of the desired plastid-modified DNA molecules with regard to a characteristic - for example with a recombination sequence - is at least 50%, preferably at least 70%, entirely is particularly preferably at least 90%, most preferably at least 95% of the totality of all plastid DNA molecules in a plant or a tissue, cell or plastid thereof. Predominantly homoplastic or predominantly homotransplastome plants can be converted into homoplastome or homotransplastome plants by further maintaining the selection pressure and, if necessary, repetitive regenerations. In a particular embodiment, therefore, a predominantly homoplastic or homotransplastome plant is purely homoplastic or homotransplastoma. A plant which, for example, is predominantly homoplastom or homotransplastom or purely homoplastom or homotransplastom with regard to a recombination sequence is consequently referred to as a "master plant". The proportion of the desired plastid-modified DNA molecules with regard to a feature can be determined, for example, by Southern analysis - as described by way of example in Example 4 - in the manner known to the person skilled in the art. The relationship between the plastidic starting DNA molecules and the plastidic DNA molecules modified with regard to a feature can be determined by comparing the intensity of the respective bands.

"Plant organism or cells derived therefrom" generally means any cell, tissue, part or reproductive material (such as seeds or fruits) of an organism which is capable of photosynthesis. Included in the scope of the invention are all genera and species of higher and lower plants in the plant kingdom. Annual, perennial, monocot and dicot plants are preferred. Included are mature plants, seeds, shoots and seedlings, as well as parts derived from them, propagation material (for example tubers, seeds or fruits) and cultures, for example row or callus cultures. Mature plants mean plants at any stage of development beyond the seedling. Seedling means a young, immature plant at an early stage of development. Preferred are plants the following plant families: Amaranthaceae, Asteraceae, Brassicaceae, Carophyllaceae, Chenopodiaceae, Compositae, Cruciferae, Cucurbitaceae, Labiatae, Leguminosae, Papilionoideae, Liliaceae, Linaceae, Malvaceae, Rosaceae, Rubiaceae, Saxifragaceae, Scrophulariaceae, Solanacea, Sterculiaceae, Tetragoniacea, Theaceae, Umbelliferae.

Preferred monocotyledonous plants are selected in particular from the monocotyledonous crop plants, such as, for example, the family of the Gramineae such as rice, maize, wheat or other types of cereals such as barley, millet, rye, triticale or oats, and sugar cane and all types of grasses.

Preferred dicotyledonous plants are in particular selected from the dicotyledonous crop plants, such as, for example

- Asteraceae such as sunflower, tagetes or calendula and others,

- Compositae, especially the genus Lactuca, especially the species sativa (salad) and others,

- Cruciferae, especially the genus Brassica, especially the species napus (rape), campestris (turnip), oleracea cv Tastie (cabbage), oleracea cv Snowball Y (cauliflower) and oleracea cv Emperor (broccoli) and other types of cabbage; and the genus Arabidopsis, especially the species thaliana as well as cress or canola and others,

- Cucurbitaceae such as melon, pumpkin or zucchini and others,

- Leguminosae especially the genus Glycine, especially the type max (soybean) soy, as well as alfalfa, peas, beans or peanuts and others

Rubiaceae, preferably of the subclass Lamiidae, such as, for example, Coffea arabica or Coffea liberica (coffee shrub) and others,

- Solanaceae especially the genus Lycopersicon, especially the species esculentum (tomato) and the genus Solanu, especially the species tuberosum (potato) and melongena (eggplant) as well as tobacco or peppers and others,

Sterculiaceae, preferably of the subclass Dilleniidae such as Theobroma cacao (cocoa bush) and others, Theaceae, preferably of the subclass Dilleniidae, such as, for example, Camellia sinensis or Thea sinensis (tea bush) and others,

i - Umbelliferae, especially the genus Daucus (especially the species carota (carrot)) and Apium (especially the species graveolens dulce (Seiarie)) and others more; and the genus Capsicum, especially the species annum (pepper) and others,

as well as flax, soybeans, cotton, hemp, flax, cucumber, spinach, carrot, sugar beet and the various types of trees, nuts and wines, in particular banana and kiwi.

Also includes ornamental plants, useful or ornamental trees,

Flowers, cut flowers, shrubs or lawn. Examples include, but are not limited to, angiosperms, bryophytes such as hepaticae (liverwort) and musci (mosses); Pteridophytes such as ferns, horsetail and lycopods; Gymno sperms such as conifers, cycads, ginkgo and gnetals, the families of rosaceae such as rose, ericaceae such as rhododendrons and azaleas, euphorbiaceae such as poinsettias and croton, caryophyllaceae such as carnations, solanaceae such as petunias, Gesneriaceae such as the Usambareae, like the usamamineae, like the usambareaceae like orchids, iridaceae like gladiolus, iris, freesia and crocus, compositae like marigold, geraniaceae like geranium, liliaceae like the dragon tree, moraceae like ficus, araceae like philodendron and others.

Plant organisms in the sense of the invention are further photosynthetically active capable organisms, such as algae, cyanobacteria and mosses. Preferred algae are green algae, such as algae of the genus Haematococcus, Phaedactylum tricornatum, Volvox or Dunaliella. Synechocystis is particularly preferred. Cyanobacteria such as Synechocystis are photosynthetically active organisms that, although they do not contain plastids, are quasi plastids and - like the plastids - contain numerous copies of their genomic DNA. In addition, they are - from a developmental biological point of view - the forerunners of today's plastids (Bauer J et al. (2001) Cell Mol Life Sei 58 (3): 420-33).

Most preferred are Arabidopsis thaliana, Nicotiana tabacum, Tagetes and Brassica napus as well as all genera and species that are used as food or feed, such as the cereals described, or for the production of oils such as oilseeds, types of nuts, soy, sunflower, pumpkin and peanut.

In the context of this invention, “recombinase” means sequence-specific recombinase. Sequence-specific recombinases comprise proteins which can catalyze a reciprocal exchange of DNA double strands between two DNA segments, the respective recombinase having a preference for DNA segments with a specific nucleic acid sequence. Said recombinant DNA sequences are not necessarily identical.

The recombinases used in the context of this invention are used in particular for the integration of DNA sequences. If other factors are required for this process - in addition to the recombinase itself - the complex of recombinase and the corresponding integration factors itself is also to be understood as a recombinase in the sense of the invention. For example, integration using the λ recombinase is only possible efficiently if the IHF (“Integration host factor”) is also present. In a preferred embodiment, said comprises

Recombinase complex no factors that favor cutting out previously by recombining inserted DNA sequences. For example, the complex of λ recombinase and IHF does not additionally include the factor xis.

The speed of a recombination reaction of a sequence-specific recombinase using its preferred recognition sequence (for example its natural recognition sequence) is preferably at least ten times as high as for the use of any other non-homologous sequence, preferably at least one hundred times as high, particularly preferably at least one thousand times as high , most preferably at least ten thousand times as high. A “non-homologous sequence” preferably means those which, with the same length, have a sequence identity to the preferred recognition sequence (for example their natural recognition sequence) of the respective recombinase of less than 90%, preferably less than 50%, particularly preferably less than 30% exhibit.

Recombinases which alone catalyze an almost irreversible integration are generally preferred. These preferably come from organisms that normally live at temperatures between 10 and 45 ° C. Although most of the known recombinases have been isolated from prokaryotes, bacteriophages or viruses, the invention is not limited to these, but can also advantageously use recombinases from eukaryotic organisms. The sequence-specific recombination used in the context of this invention is not limited to a specific mechanism. As a rule, however, the recombination takes place in two steps: a sequence-specific cut of the recognition sequence followed by a relinking, whereby there is usually a covalent protein-DNA intermediate between the target sequence and the recombinase. For some recombinations, only the recombinase as a protein is required. Others may need additional help factors. The recombination system used is preferably independent of the enzyme equipment of the recipient plastid.

Recombinases belonging to the families of integrases or resolvases / invertases are particularly preferred (Stark WM et al.

15 (1992) Trends Genet 8: 432-439). Integrases (tyrosine recombinases) have a tyrosine in the active center, which acts as a nucleophile in the cleavage step and is involved in the formation of a 3'phosphotyrosyl bond with the DNA. The reaction takes place via a "Holliday Intermediate". Resolvases / invertases (serine

20 recombinases) have a serine in the active center, which acts as a nucleophile in the cleavage step and is involved in the formation of a 5 'phosphoserinyl bond with the DNA. The reaction does not take place via a "Holliday Intermediate".

The family of integrases (also called tyrosine recombinases) include, for example, Cre recombinase from bacteriophage PI, FLP recombinase from yeast, λ integrase from phage Lamda and R recombinase from pSRI plasmid from yeast Zygosaccharomyces rouxii (Araki et al (1985) J Mol Biol.

30 182: 191-203) and the h recombinase (Argos, et al. (1986) EMBO J 5: 433-440, 1986; Sadowski PD (1995) Progress Nucl Acid Res Mol Biol 51: 53-91).

The λ integrase usually requires further factors that bind to 35 sequences adjacent to the immediate recognition sequence. In addition to the recombinase (also called Int for integrase), IHF ("Integration host factor") is also required. The factor xis is additionally required for the reverse reaction, the cutting out. 40

The resolvase / invertase family includes, for example, the ΦC31 recombinase, R4 recombinase (GenBank Acc.-No .: D38173 nucleotides 292 to 1701), the Hin inversion system from Salmonella typhimurium and the TP901-1 recombinase (Hallet and Sherratt 45 (1997) FEMS Microbiol Rev 21: 157-178). Other family members include the inverted gin of Mu phage, one of bacteriophage P and pin of el4 phage.

In contrast to the integrase family, the members of the resolvases generally have a conserved N-terminal catalytic domain (Crellin and Rood (1997) J Bacteriology 15 179 (16): 5148-5156; Christiansen et al. (1996) J Bacteriology 178 (17): 5164-58173). Like some of the Cre-type recombinases, some resolvases are independent of host factors (Thorpe and Smith (1998) PNAS USA 95: 5505-5510). The resolvases / invertases use a different reaction mechanism than the integrases (Gopaul DN and van Duyne GD (1999) Curr Opin Structural Biol 9: 14-20; Jayaram M (1994) TIBS 19: 78-83). The subgroup of “extended resolvases” is particularly preferred (Brondsted L and Hammer K (1999) Appl Environm Microbiol 65: 752-758). These have homology to the other resolvases / invertases, but have an extension at the C-terminus compared to these. The advantage of these recombinases is that they require little to no helper factors to catalyze the recombination between two recognition regions (Brondsted L and Hammer K (1999) Appl Environm Microbiol 65: 752-758). However, additional factors are required for the efficient catalysis of the reverse reaction. Expression of the corresponding recombinase alone can therefore, with suitable positioning of the recognition sequences, bring about an almost irreversible integration of a heterologous DNA sequence under these conditions. The recombinases of phages TP901-1 and ΦC31 from this subgroup are particularly preferred.

The recombinases of the integrase family are preferably used because they naturally mediate intermolecular recombinations as a rule.

In the present invention, particular preference is given to using recombinases which recombine two different recognition regions and which the recombinase no longer recognizes and uses the hybrid recognition region which arises during the recombination efficiently or only with a further factor.

While extensive sequence homology is required for the recombination events of the homologous recombination (eg double crossovers), the recognition sequence-specific recombination mediates DNA rearrangements between segments which have no extensive homology. The recombination events occur in the recognition sequence-specific recombination only at specific locations of the DNA (Craig NL (1988) Annu Rev Genet 22: 77-105). Recombination systems such as the λ integrase from the λ phage, the recombinase from the mycobacteriophage L5 (Pena CEA et al. (2000) Proc Natl Acad Sei USA 97: 7760-7765; Lee MH et al. (1991) Proc Natl Acad Sei USA 88: 3111-3115), the recombinase from the phage ΦC31 (Kuhstoss S and Rao RN (1991) J Mol Biol 222: 897-908; Groth AC et al. (2000) Proc Natl Acad Sei USA 97: 5995- 6000; Rausch H and Lehmann M (1991) Nucl Acids Res. 19: 5187-5189; Thorpe HM and Smith MCM (1998) Proc Natl Acad Sei USA 95: 5505-5510; Kuhstoss S. et al. (1991) Gene 102: 145 -146; Harris JE et al. (1983) Gene 22: 167-174) or the phage recombinase

TP901-1 (Brondsted L and Hammer K (1999) Appl Environm Microbiol 65: 752-758; Bruener A et al. (1999) J Bacteriol 181: 7291- 7297; Koch B et al. (1997) Appl Environm Microbiol 63: 2439-2441, Christiansen B et al. (1996) J Bacteriol 178: 5164-5173) naturally use two recognition regions with different sequences in their sequence, usually called attP and attB, for a recombination event. These are recombined into new, hybrid recognition regions (attR and attL) that are not used efficiently by the corresponding recombinase alone. Only when one or more other factors are added, these recognition regions are recognized again and a recombination (back reaction) between them can take place. A large number of factors which play a role in the catalytic action of the λ recombinase are known (Landy A (1993) Curr Opin Gen Develop 3: 699-707). Other preferred recombinases to be mentioned are: xisF from Anabaena (Ramaswamy KS et al. (1997) Mol Microbiol 23: 1241-1249; Carrasco CD et al. (1994) Genes Develop 8: 74-83), phage integrase ΦLC3 (Lillehaug D (1997) Gene 188: 129-136), recombinase encoded by the sre gene of the R4 phage (Matsuura M et al. (1996) J Bacteriol 178: 3374-3376).

In addition to the preferred recombinase systems that naturally mediate an irreversible reaction, the known recombinase systems such as Cre / lox; FLP / frt from yeast; R-RS from Zygosaccharomyces rouxii (Onouchi, H et al., 1991, Nucl. Acids Res. 19: 6373-6378) or Gin-gix from the bacteriophage Mu (Maeser S and Kahmann R (1991) Mol Gen Genet 230 : 170-176) can be used in the context of this invention. These systems naturally mediate a reversible recombination, whereby the balance is usually shifted to the side of the excision. For use in the context of this invention, these systems are preferably made almost irreversible (see below).

The recombinases from ΦC31 and TP901-1 are used with particular preference. Most preferably, the recombinases with nucleic acid sequences are deposited under GenBank Acc.-No Y14232 (TP901-1; complementary bp 30 to 1487) or X59938 (ΦC31; nucleotides 232 to 2073).

Various methods are conceivable to introduce a recombinase into plastids or to express them there. Examples include, but are not limited to:

a) Nuclear expression using plastid transit peptides

An expression cassette coding for a recombinase can be constructed in the manner known to the person skilled in the art, inserted into the cell nucleus and - optionally - stably integrated into the chromosomal DNA. The expression is then transient or - when integrated into the chromosomal DNA - stable. For transport into the plastids, the recombinase is preferably expressed in fusion with a plastid localization sequence (PLS). Methods of specifically transporting proteins that are not per se localized in the plastids into the plastids as well as various PLS sequences are described (Klosgen RB and Weil JH (1991) Mol Gen Genet 225 (2): 297-304; Van Breuse- gem F et al. (1998) Plant Mol Biol 38 (3): 491-496). Preferred are PLS which are enzymatically split off from the recombinase part after translocation of the recombinase into the plastids. Particularly preferred is the PLS, which is derived from the plastid Nicotiana tabacum transketolase or another transit peptide (e.g. the transit peptide of the small subunit of Rubisco or the ferredoxin NADP oxidoreductase as well as the isopentenyl pyrophosphate isomerase-2) or its functional equivalent. In principle, all promoters which allow expression in plants are suitable for nuclear expression. Examples are given below. Constitutive promoters such as the 35S promoter of the CaMV or the nitrilase-1 promoter of the nitl gene from Arabidopsis (GenBank Acc.-No .: Y07648.2, nucleotides 2456 to 4340; Hillebrand H et al. (1998) Plant Mol Biol 36 (l): 89-99; Hillebrand H et al. (1996) Gene 170 (2): 197-200).

Preferred PLS sequences are:

i) the transit peptide of isopentenyl isomerase (IPP) from Arabidopsis (GenBank Acc. -No.: NC_003074; nucleotides 604657 - 604486) ii) Transit peptides derived from the small subunit (SSU) of ribulose bisphosphate carboxylase (Rubisco ssu) from, for example, pea, corn, sunflower or Arabidopsis.

Arabidopsis thaliana: GenBank Acc.-No .: e.g. AY054581, AY054552;

Pea, GenBank Acc. -No. : e.g. X00806, nucleotides 1086 to 1256; X04334, X04333 (Hand JM (1989) EMBO J

8 (11): 3195-206). Particularly preferred here: expression cassette and transit peptide (pea, rbcS3A) from vector pJIT117 (Guerineau F (1988) Nucleic Acids Res 16 (23): 11380. The peptide sequence according to SEQ ID NO: 11 is particularly preferred. Most preferred for the The nucleic acid sequence according to SEQ ID NO: 10 is used for the construction of corresponding expression constructs.

- Maize, GenBank Acc. -No. : e.g. S42568, S42508

Sunflower, GenBank Acc. -No. : Y00431, nucleotides 301 to 465.

iü) Transit peptides derived from genes of vegetable fat biosynthesis such as the transit peptide of the plastid "Acyl Carrier Protein" (ACP) (e.g. Arabidopsis thaliana beta-ketoacyl-ACP synthetase 2; GenBank Acc.- No.: AF318307), the stearyl-ACP- Desaturase, β-ketoacyl-ACP synthase or the acyl-ACP thioesterase (e.g.

A. thaliana mRNA for acyl- (acyl carrier protein) thioesterase: GenBank Acc. -No. : Z36911).

iv) the transit peptide of GBSSI ("Starch Granule Bound Synthase I")

v) the transit peptide of the LHCP II genes.

The plastid transketolase from tobacco (SEQ ID NO: 12) is particularly preferred. Various PLS nucleic acid cassettes can be used in the three reading frames as Kpnl / BamHI fragments for the expression of corresponding fusion proteins (the translation start (ATG codon) is located in the Ncol interface) (pTP09 SEQ ID NO: 13; pTPlO SEQ ID NO: 14; pTPll SEQ ID NO: 15). Another example of a PLS that can be used advantageously is the transit peptide of the plastidic isopentenyl pyrophosphate isomerase-2 (IPP-2) from Arabidopsis thaliana (SEQ ID NO: 16). The nucleic acid sequences coding for three cassettes (corresponding to the three reading frames) of the PLS of the isopentenyl pyrophosphate isomerase-2 (IPP-2) from Arabidopsis thaliana (EcoRV / Sall cassettes with ATG; IPP-9 SEQ ID NO: 17) can be used with very particular preference ; IPP-10 SEQ ID NO: 18; IPP-11 SEQ ID NO: 19).

The nucleic acids according to the invention can be produced synthetically or obtained naturally or contain a mixture of synthetic and natural nucleic acid constituents, and can consist of different heterologous gene segments from different organisms.

The sequence coding for the transit peptide can comprise all or part of the peptide sequence of the original protein. An exact determination of the amino acid residues essential for the transport is not necessary as long as the functionality of the PLS - namely the transport into the plastids - is guaranteed and the function of the recombinase is not completely destroyed. The following PLS sequences are very particularly preferred:

PLS1: N-MASSSSLTLSQAILSRSVPRHGSASSSQLSPSSLTFSGLKSNPNITTSRRR TPSSAAAAAWRSPAIRASAATETIEKTETAGS-C (SEQ ID NO: 12). Corresponds to the PLS of plastid transketolase from Taba.

PLS2: N-MSASSLFNLPLIRLRSLALSSSFSSFRFAHRPLSSISPRKLPNFRAFSGTA MTDTKDGSRVDM-C (SEQ ID NO: 16). Corresponds to the PLS of the isopentenyl pyrophosphate isomerase-2 (IPP-2), the last methionine preferably being the start methionine of the recombinase.

In the context of this invention, fusion proteins from PLS and recombinase are subsumed under the term recombinase. If a recombinase is expressed nuclear, the recombinase is preferably understood to mean a fusion protein composed of PLS and the recombinase.

Expression in plastids

Expression in plastids can also be achieved by the direct introduction of an expression cassette for the recombinase in plastids, possibly integration into the plastid DNA and Expression of the recombinase take place. In a preferred embodiment, this expression cassette is contained in the transformation construct comprising the insertion sequence. However, expression can also be carried out from a separate expression cassette.

On the one hand, specific plastid or chromoplast promoters - as described in detail below - can be used. A targeted plastid expression can also be achieved if, for example, a viral, bacterial or bacteriophage promoter is used, the resulting expression cassette is inserted into the plastid DNA and the expression is then induced by the corresponding viral, bacterial or bacteriophage RNA polymerase. The corresponding RNA polymerase can in turn be introduced into the plastids in various ways - preferably by nuclear transformation as a fusion protein with a PLS. Corresponding methods are described (WO 95/16783, WO 97/06250, US 5,925,806). The introduction into plastids is preferably carried out by means of microinjection and particularly preferably by means of particle bombardment.

c) Introduction as RNA

The recombinase can also be introduced by the for

Recombinase-encoding mRNA, e.g., generated in vitro, e.g. can be introduced into plastids via microinjection, particle bombardment (bio- listic processes), polyethylene glycol or liposome-mediated transfection. This embodiment is advantageous since no sequences coding for the recombinase remain in the piastome or genome. The RNA encoding the recombinase is preferably produced by in vitro transcription in a manner known to the person skilled in the art.

d) Introduction as a protein

The recombinase can be introduced directly into plastids, for example via microinjection, particle bombardment (biolistic method), polyethylene glycol transfection or liposome-mediated transfection. This embodiment is advantageous since no sequences coding for the recombinase remain in the piastome or genome. A corresponding method is described for example by Segal DJ et al. (1995) Proc Natl Acad Sei USA 92: 806-810. The recombinase can alternatively be introduced into plant cells as a fusion protein with the VirE2 or VirF protein of an agrobacterium and a PLS. Corresponding methods are described, for example, for Cre recombinase (Vergunst AC et al. (2000) Science 290: 979-982). This embodiment is advantageous since no sequences encoding the recombinase remain in the genome.

Combinations of the options described above are of course also conceivable.

The expression cassette for the recombinase is preferably contained on the insertion sequence or the transformation construct. It is also possible to produce master plants which comprise an expression cassette for a recombinase stably integrated into the plastid DNA or the nuclear DNA.

The recombinase is preferably functionally present at the moment the transformation construct or insertion sequence is introduced into the plastids of the master plant.

The recombinase is particularly preferably introduced or activated simultaneously with or after the introduction of the insertion sequence into the plastids. Expression or activation at the right place and at the right time can be ensured by different approaches:

a) Inducible expression

Expression of a recombinase can be controlled using an inducible promoter, preferably a chemically inducible promoter. For this purpose, for example, the expression cassette coding for the recombinase can be stably transformed into the plastidic or nuclear DNA of a master plant. Does the transformation into that

Nuclear genome, as described above, the subcellular localization must be ensured by suitable PLS transit peptides. Shortly before or during the transformation with the insertion sequence or the transformation construct, depending on the chosen inducible one

System switched on the expression of the recombinase by application of the appropriate inductor. Various methods or promoters for induced expression are known to the person skilled in the art. Chemical substances or physical stimuli such as increased stimuli can be used as the stimulus Temperature or wound etc. act. Various examples are described below.

b) Inducible activity

The recombinase can already be present in the plastids of the master plant if the activity is only induced by suitable techniques at the selected point in time, for example by adding chemical substances. Appropriate methods have been described for sequence-specific recombinases (Angrand PO et al. (1998) Nucl Acids Res 26 (13): 3263-3269; Logie C and Stewart AF (1995) Proc Natl Acad Sei USA 92 (13): 5940-5944; Imai T et al. (2001) Proc Natl Acad Sei USA 98 (1): 224-228). In these methods, fusion proteins from the recombinase and the ligand binding domain of a steroid hormone receptor (e.g. the human androgen receptor, or mutated variants of the human estrogen receptor as described there) are used. Induction can be carried out using ligands such as estradiol, dexamethasone, 4-hydroxytamoxifene or raloxifene.

Most recombinases are active as multimers (often as tetramers) (Stark et al., 1992, TIG 8: 432-439). It is conceivable to make the multimerization inducible by, for example, exchanging the natural multimerization domains for the binding domain of a low molecular weight ligand. The addition of a dimeric ligand then causes the fusion protein to dimerize. Corresponding inducible dimerization processes and the preparation of the dimeric ligands have been described (Amara JF et al. (1997) Proc Natl Acad Sei USA 94 (20): 10618-1623; Muthuswamy SK et al. (1999) Mol Cell Biol 19 (10) : 6845-6857; Schultz LW and Clardy J (1998) Bioorg Med Chem Lett 8 (l): 1-6; Keenan T et al. (1998) Bioorg Med Chem. 6 (8): 1309-1335).

c) Cotransfection

The expression cassette coding for the recombinase is preferably introduced into the plastids simultaneously with the insertion sequence. The expression cassette for the recombinase and the insertion sequence can be present on one DNA molecule or on two separate ones. The two sequences are preferably present together on a DNA molecule, so that the expression cassette is contained in the transformation construct comprising the insertion sequence. In a particularly preferred embodiment, after homoplastomic plants have been regenerated, the sequence coding for the recombinase is removed from the piastome. Various methods are known to the person skilled in the art for this purpose, which are described in detail below.

Another object of the invention relates to fusion proteins of recombinases of the resolvase / invertase family with plastid localization sequences, and the nucleic acid sequences coding therefor. Also included are expression cassettes which contain said nucleic acid sequences under the control of a promoter which is functional in the plant cell nucleus. Corresponding promoters are known to the person skilled in the art and are described further below. Preferred resolvases / invertases are the ΦC31 recombinase, the R4 recombinase, the Hin inversion system from Salmonella typhimurium and the TP901-1 recombinase, as well as the recombinases Gin from Mu phage, Cin from bacteriophage P and pin from el4 phage. The ΦC31 recombinase and the TP901-1 recombinase are very particularly preferred. The proteins according to SEQ ID NO: 41, 42, 46, 47 and 61 are most preferred. The sequences according to SEQ ID NO: 40 and 45 are particularly preferred as expression cassettes. The sequence of the TP901-1 recombinase according to SEQ ID NO: 46 and 47 contains an amino acid exchange compared to the one stored in GenBank (SEQ ID NO: 61).

The invention further relates to expression cassettes which contain nucleic acid sequences coding for recombinases of the resolvase / invertase family under the control of a promoter which is functional in plant plastids. Corresponding promoters are known to the person skilled in the art and are described further below. Preferred resolvases / invertases are the ΦC31 recombinase, the R4 recombinase, the Hin inversion system from Salmonella typhimurium and the TP901-1 recombinase, as well as the recombinases gin from Mu phage, Cin from bacteriophage P and pin from el4 phage. The ΦC31 recombinase and the TP901-1 recombinase are very particularly preferred.

"Recombinase recognition sequence" (as a result of "RE sequence") generally means those sequences which are under the conditions in the

Plastids of the plant cell or plant used in each case allow recognition and subsequent use as a recombination substrate by a recombinase.

The speed of a recombination reaction of a specific recombinase using a RE sequence (for example its natural recognition sequence) is preferably at least ten. times as high as for the use of any other non-homologous sequence, preferably at least a hundred times as high, particularly preferably at least one thousand times as high, most preferably at least ten thousand times as high. Here, a “non-homologous sequence” preferably means those that are the same

Length, a sequence identity to the preferred recognition sequence (for example its natural recognition sequence) of the respective recombinase of less than 90%, preferably less than 50%, particularly preferably less than 30%.

RE sequences of the sequence-specific recombinases described above are particularly preferred. Preferably, the RE sequence of a particular recombinase is singular in the plastid DNA, i.e. a recombination is only created at the predefined location. However, cases are also conceivable in which there is more than one RE sequence in the piastom. This is particularly the case if the RE sequence is located in duplicated genes (e.g. in inverted "repeats"). In the latter case, there is more than one identical RE sequence, but their context is identical, so that a targeted insertion is also carried out here. It is even preferred that the integration be carried out in all copies, so that an insertion in all copies is also necessary. RE sequences that occur more than once in a piastom, but are localized in the same plastomic context (e.g. in "repeats" or gene duplications) are subsumed within the scope of this invention under the term of the singular RE sequences. Of course, singular RE sequences from different recombinases can be present in parallel in one piastome.

Instead of the natural RE sequences of the recombinases used, modified recognition regions can also be used, as described, for example, for the FLP or Cre recombinase (Senecoff JF and Cox MM (1986) J Biol Chem 261: 7380-7386; Albert H et al (1995) Plant J 7: 649-659). However, the natural sequences are preferably used in the context of this invention. In addition, the recombinases themselves can be improved by mutagenesis, as has been shown for the FLP recombinase (Buchholz F et al. (1998) Nature Biotech 16: 657-662).

It is also conceivable to use two RE sequences that do not recombine with one another (which are recognized by one and the same or different recombinases) to flank a DNA segment to be integrated. These can preferably be recombined with compatible recognition regions in another molecule (Hoess RH et al. (1986) Nucl Acids Res 14: 2287-2300). A basic distinction can be made between "wild-type" RE sequences and "pseudo" RE sequences. Here, "wild-type" RE sequence means sequences such as those used by the recombinase in the natural system. For example, the loxP sequence for the cre recombinase or the FRT sequence for the FLP recombinase or the GIX sequence for the gin recombinase. These sequences can be derived from their homologous origin systems (for example the natural phages) and can advantageously be used in the context of this invention. "Pseudo" RE sequences means those sequences which can be recognized by a recombinase and used as a substrate for recombination, but which have sequences which are not identical to the "wild-type" RE sequence. Such sequences can be isolated from heterologous organisms, for example, or generated by means of artificial mutagenesis. For example, some recombinases - such as ΦC31 - can cause recombinations in the genomes of some eukaryotes, so they also have functional RE sequences here that are similar but not identical to the "wild-type" RE sequence. "Pseudo" RE sequences can be identified by means of sequence alignments, secondary structure comparison, deletion or point mutation analysis. Functional tests are advantageously used to determine the effect of the recombinase on the RE sequences. Also included are hybrid RE sequences which are composed of parts of individual RE sequences - for example - each of a portion of a "wild-type" RE sequence and a "pseudo" Re sequence.

The recombinase is introduced into the plastids before, at the same time and / or after the introduction of the transformation construct.

Preferably, the plant used or the cell derived from it is predominantly homoplastomic or homotransplastomic with respect to the RE sequence, i.e. that the majority of the plastid DNA molecules contained in a plastid have this RE sequence. Such plants are also referred to as master plants in the context of this invention.

In principle, two types of RE sequences can be used:

a) Natural, endogenous RE sequence

A sequence that occurs naturally in the piastome can act as a recognition sequence for a chimeric recombinase, for example. Chimeric recombinases can consist, for example, of a zinc finger domain binding the sequence of said recognition region and a catalytic domain of a recombinase (WO 96/06166). Like such zinc finger domains or such chimeric recombinases are known to the person skilled in the art (Beerli RR et al. (2000) Proc Natl Acad Sei USA 97 (4): 1495-1500; Beerli RR et al. (2000) J Biol Chem 275 (42) : 32617-32627; Segal DJ and Barbas CF 3rd. (2000) Curr Opin Chem Biol 4 (l): 34-39; Kang JS and Kim JS (2000) J Biol Chem 275 (12): 8742-8748; Beerli RR et al. (1998) Proc Natl Acad Sei USA 95 (25): 14628-14633; Kim JS et al. (1997) Proc Natl Acad Sei USA 94 (8): 3616-3620; Klug A (1999) J Mol Biol 293 (2): 215-218; Tsai SY et al. (1998) Adv Drug Deliv Rev 30 (1-3): 23-31; Mapp AK et al. (2000) Proc Natl Acad Sei USA 97 (8): 3930-3935; Sharrocks AD et al. (1997) Int J Biochem Cell Biol 29 (12): 1371-1387; Zhang L et al. (2000) J Biol Chem 275 (43): 33850-33860). Methods for the production and selection of zinc finger DNA binding domains with high sequence specificity are described (WO 96/06166, WO 98/53059, WO 98/53057).

In addition, it is conceivable to subject known recombinases to mutagenesis until they recognize a selected sequence of the plastid genome of a plant species under consideration as a substrate. Plants with endogenous, natural RE sequences are quasi-naturally occurring "master plants". The RE sequence is naturally homoplastic in them. This eliminates the need to introduce and select artificial RE sequences.

Artificially introduced RE sequence

The person skilled in the art is aware that the RE sequence introduced into a master plant need not be natural. In principle, any RE sequence - for example a natural, mutated or pseudo RE sequence - of any recombinase can be inserted at any location on the plastid DNA. The preparation is preferably carried out using a construct for inserting the RE sequence (as a result of the RE construct).

The RE construct preferably comprises a selection marker in order to facilitate the selection of transplastomeric plants with the successfully inserted RE sequence, which is required to produce corresponding master plants. Various selection markers are known to the person skilled in the art which enable the selection of plastids (see below). Preferred are aadA, nptll or BADH, with aadA being particularly preferred. The selection is carried out, for example, using the “segregation and sorting” process known to the person skilled in the art (described by way of example in Example 4). The selection marker is preferably constructed such that a subsequent deletion from the piastome is made possible. Corresponding methods are known to the person skilled in the art and are described below.

In addition to the selection marker, the RE construct can contain further sequences. These can contain, for example, further regulatory elements for the expression of the insertion sequences to be introduced as a result. In a preferred embodiment, the selection marker introduced in the context of the construct for inserting the RE sequence is deleted after the homoplastic master plant has been obtained by methods known to the person skilled in the art (see below).

Alternatively, natural regulatory elements present endogenously at the insertion site can also be used for the expression of genes to be introduced.

In a preferred embodiment, the RE construct comprises further flanking sequences (A and B) which enable a site-specific insertion on at least one, preferably on both sides of the RE sequence and which have a sufficient length and homology to corresponding target sequences in the piastome (A 'and B ') have to ensure a site-specific insertion by means of homologous recombination.

RE sequences not naturally occurring in the plastid DNA can be introduced into the plastid DNA in various ways. Examples include:

a) Integration using homologous recombination (e.g. double crossover)

The integration into the plastid genome is preferably carried out with the aid of the above-described methods of homologous recombination (e.g. double crossover) which are well known to the person skilled in the art.

b) Integration using natural or endogenous recognition sequences for an enzyme suitable for inducing DNA double-strand breaks (as a result of the DSB recognition sequence; see below).

c) Integration as a result of recombination with recombinases. The method according to the invention itself can be used, for example, to integrate further RE sequences. Different locations of the localization or integration of a RE sequence (in the case of endogenous RE sequences already present) or integration (in the case of artificially generated RE sequences) are possible for the RE sequence. Examples include:

a) Localization (integration) in a transcriptionally silent region

Localization (integration) of the RE sequence in a transcriptionally silent region of the plastid genome (intergenic region) is the preferred embodiment. A disruption of the plastid functions can be largely excluded. It should be noted here that for an expression (e.g. of selection markers or other genes of interest) appropriate regulatory elements such as promoters etc. may have to be included.

b) Localization (integration) in a transcriptionally active, but non-coding (intercistronic) region

This localization (integration) has the advantage that the insertion sequence to be introduced is ultimately encoded in a plastid operon and the promoter (s) or terminator (s) do not have to be introduced separately, but the endogenous ones present at this locus can be used ( but do not have to). In such a case, only ribosome binding sites should be present at a suitable distance upstream of the coding region of the foreign genes to be introduced.

However, it is also conceivable that an intergenic region is not completely transcriptionally silent because, for example, the transcription of an adjacent gene or operon is terminated only inefficiently.

c) Localization (integration) in a transcriptionally active coding region.

The localization (integration) of the RE sequence at a non-coding locus described under a) and b) has the

The advantage that the insertion of the foreign DNA is not likely to influence the function of the plastid genome. However, non-coding areas are less conserved than coding areas. In order to have a method which is as universal as possible and which works in many plant species, in a particularly preferred embodiment the RE sequence (and consequently the insertion sequence) is in the coding sequence of an existing gene. The destruction of the gene function by the introduction of the RE sequence (in the case of an artificially generated RE sequence) or by the subsequent introduction of the insertion sequence is prevented in an inventive manner in that the RE sequence or the insertion sequence in a preferred variant of this embodiment in An intron is inserted. In this way, the complete coding mRNA of the gene into which the RE sequence has been inserted is generated again by splicing the pre-RNA. Alternatively, the insertion can also take place in a non-essential plastid gene. The gene destroyed by the insertion of the RE sequence can also be reintroduced in a subsequent transformation.

The RE construct preferably has - particularly if the insertion into a transcriptionally active or even translated piastome region - the structure and sequence of an intron. As a rule, naturally occurring introns are modified so that they meet the requirements of the method according to the invention. The insertion is preferably carried out in such a way that the inserted sequence is completely removed by splicing the pre-mRNA. The spliced RNA (ie the artificial intron) then represents the mRNA, for example for translation on its encoded proteins. Since the transcription of the intron is subject to the regulatory control of the gene into which the intron has been integrated, all regulatory elements can be viewed upstream or downstream of the genes or gene of interest in the intron. This allows you to keep the constructs small. In addition, all introns can be used if the relevant factors that mediate the splicing are simultaneously expressed in the plastids or imported into them. The factors supporting the splicing are preferably coded in the intron itself. In this embodiment, group II introns are particularly preferred which themselves encode at least one of the factors necessary for the splicing. This includes the Ll.ltrB intron from Lactococcus. Also preferred are introns that occur naturally in plastids of higher plants, especially introns of group II, very particularly preferably introns that code for a protein, most preferably introns of the trnK genes of the plastid genome. In the latter case, the introns from the trnK genes of the plastids from the Arabidopsis, maize and tobacco species are particularly preferred.

Preferred are introns with a self-splicing activity that does not depend on other protein factors or use the general factors that are universal and thus also in Plastids are present. These introns include, for example

a) the intron of Group I from Tetahymena (GenBank Acc.-No .: X54512; Kruger K et al. (1982) Cell 31: 147-157; Roman J and Woodson SA (1998) Proc Natl Acad Sei USA 95: 2134 -2139)

b) the group II rll intron from Scenedesmus obliquus (GenBank Acc.-No .: X17375.2 nucleotides 28831 to 29438; Holländer V and Kück U (1999) Nucl Acids Res 27: 2339-2344; Herdenberger F et al. (1994) Nucl Acids Res 22: 2869-2875; Kück U et al. (1990) Nucl Acids Res 18: 2691-2697).

c) the Ll.LtrB intron (GenBank Acc.-No .: U50902 nucleotides 2854 to 5345)

d) the trnK intron from Arabidopsis (GenBank Acc.-No .: AP000423 neucleotides complementary 1752 to 4310), maize (GenBank Acc.-No .: X86563 nucleotides complementary 1421 to 3909) or tobacco (GenBank Acc.-No .: Z00044 nucleotides complementary 1752 to 4310)

Both intrinsic and introns naturally occurring in the plastids of the respective plant can be used. In order to avoid instabilities caused by sequence duplication, introns of a different species - for example trnK introns of a different species - are preferred. In a preferred embodiment, introns naturally occurring in the plastids of the respective plant are modified such that they can still fulfill their function, but the sequence homology is less than 95%, preferably 80%, particularly preferably 70% of the sequence of the starting intron ,

Introns which naturally encode a DSB enzyme (in particular a homing endonuclease) are particularly preferred. The intron Cp.LSU2 from Chlamydomonas pallidostigmatica, which encodes the enzyme I-Cpal (Tunnel M et al. (1995) Mol Biol Evol 12: 533-545), is particularly preferred. Group II introns from the yeast mitochondria are also preferred.

In a preferred embodiment, the intron sequence is adapted to the insertion site so that they can splice at this locus. The adaptation can affect the internal guide sequence (IGS) for group I introns or the exon binding sequence (EBS) I or / and II for group II introns. In the case of the trnK intron from maize, it could be shown that the corresponding mRNA is edited in barley plastids (Vogel J et al. (1997) J Mol Biol 270: 179-187). In a preferred embodiment, the matK gene in the trnK intron from maize is therefore already modified by a corresponding His / Tyr exchange at the DNA level, so that RNA editing is no longer necessary.

In the case of group I introns, the splice point is determined by mating the IGS with the exon of the corresponding transcript located 5 'and / or 3' to the intron (Lambowitz AM & Beifort M (1993) Annu Rev Biochem 62: 587-622). Using techniques known to the person skilled in the art, such as PCR or the synthetic generation of nucleotide sequences, the IGS of any group I introns can be adapted accordingly such that splicing takes place at the predefined insertion point within the DSB recognition region. The intron CpLSU2 from C. pallidostigmatica, which codes for the homing endonuclease I-Cpal, is preferably used. Group I intron from Tetrahymena thermophila (GenBank Acc.-No .: V01416 J01235; nucleotides 53 to 465) is also preferred. The IGS that can be found naturally can be adapted to the new insertion site by techniques known to those skilled in the art.

Group II self-splicing introns have a conserved structure and generally consist of 6 different domains. Domain I contains the exon binding sites (EBS1 and EBS2), which interact with the exon 5 'from the intron during the splicing process. In addition, there is an interaction between the "5 region" and the "δ 'region" at the 3' exon (Lambowitz AM & Beifort M (1993) Annu Rev Biochem 62: 587-622; Michel F & Ferat JL (1995) Annu Rev Biochem 64: 435-461). These sequences can be adapted in each case by techniques known to the person skilled in the art, such as synthetic creation of the introns or suitable PCR methods, in such a way that a correct choice of the splice sites at the insertion location selected in the DSB recognition region is ensured.

Of course, several RE sequences for preferably different recombinases can also be incorporated via one RE construct. The RE construct is preferably created in such a way that no sequences are duplicated after integration into the piastome (no homologies to native sequence regions).

In this way, a plant which is predominantly homotransplastome with respect to the inserted RE sequence is preferably first generated, which has a RE sequence in all or the majority of the plastids of the plant under consideration. Such plants can advantageously be used as master plants. Even if the effort to insert an artificial RE sequence into the plastid DNA is relatively high and, in the case of insertion using homologous recombination (e.g. crossover), corresponds to the process for plastid transformation currently described in the prior art, this effort only has to be done once operate. The homotransplastome master plant obtained can then be used for any number of different subsequent transformations using the method according to the invention, which enables a significant increase in the transformation efficiency: instead of having to go through the conventional selection process for a homotransplastome plant every time, it only has to be done once here will be realized.

Due to the large number of recombinases with defined RE sequences described in the prior art, it is possible and preferred to generate master plants which have incorporated several different singular RE sequences into their plastid genome. Different transformation vectors can then also be introduced - also simultaneously. When using different recognition regions that are either already present in the master plant or that are introduced with the newly transformed transformation construct, the process can be repeated as often as required. Not all vectors introduced have to be inserted using recombinases (different insertion sequences can also be inserted using different integration mechanisms); not all - optionally none - transformation constructs have to code a selection marker. Selection markers can also be present in trans on another construct. This additional construct can also comprise only one selection marker (for example "binding type marker" such as a point mutation in the 16SrDNA to achieve spectomycin resistance).

For RE sequences to be used advantageously within the scope of the invention, the recognition sequences for the respective recombinases listed in Table 1 below may be mentioned by way of example, but not by way of limitation. Table 1: Recognition sequences of recombinases

This also includes deviations (degenerations) in the recognition sequence, which nevertheless enable recognition and recombination by the respective recombinase, for example pseudo RE sequences. Furthermore, core sequences (“core” sequences) of these recognition sequences are included. It is known that the inner parts of the recognition sequences are sufficient for a recombination and that the outer parts are not necessarily relevant, but can also determine the efficiency of the recombination. Furthermore, the specified sequences can be supplemented or expanded as desired on both sides, for example by adopting larger flanking areas from the corresponding natural sequences. In this respect, the term “RE sequence” also encompasses all essentially identical recognition sequences. Essentially identical recognition sequences mean those recognition sequences which, although they have deviations from the recognition sequence found to be optimal for the particular enzyme, still permit recombination by the same.

Since the loxP sequence has so far only been found in the PI phage genome, the use of the identical loxP sequence requires a prior introduction of the same into the piastome (see below, e.g. using homologous recombination). However, various studies have already shown that an exact identity with the natural loxP sequence is not absolutely necessary for Cre-mediated recombination (Sternberg et al. (1981) J Mol Biol 150: 487-507; Sauer J (1992) Mol Biol 223: 911-928; Sauer (1996) Nucl Acids Res 24: 4608-4613).

Orientation and localization of the RE sequences

Deletion: For a deletion, two RE sequences, which can recombine with each other through the action of a recombinase, are placed intramolecularly with the same orientation. The sequence to be deleted is between the two RE sequences. Two molecules are formed and each carries one of the two sequences that result from the recombination of said RE sequences. Insertion: The insertion is the reverse reaction of the deletion. Two RE sequences that can be recombined are placed on 2 molecules. In the simplest application, the molecule to be inserted is circular. The integration is carried out in the manner in which both RE sequences are arranged in a corresponding orientation. The resulting product then carries two sequences, which result from the recombination of said RE sequences, in a corresponding orientation.

It is also possible to place two RE sequences, which cannot recombine with one another, on a molecule with the same orientation. In the second molecule, which then does not necessarily have to be circular, two RE sequences, which cannot recombine with each other but with those on the other molecule, are then also placed with the same orientation (see "Replacement" strategy below). The result of the preferred reaction is an exchange of the sequences between the RE sequences on both molecules.

Inversion: Two RE sequences that can be recombined are placed on a molecule with opposite orientation. The associated recombinase can then cause an inversion of the sequences located between the RE sequences. If the hybrid RE sequences resulting from the recombination cannot recombine with one another by the action of said recombinase alone, the inversion is irreversible.

"Enzyme suitable for the induction of DNA double-strand breaks on the recognition sequence for the targeted induction of DNA double-strand breaks" (as a result of "DSBI enzyme" for "double strand-break inducing enzvme") generally means all those enzymes which are capable to generate sequence-specific double-strand breaks in double-stranded DNA. For example, but not restrictive:

1. Restriction endonucleases, preferably type II restriction endonucleases, particularly preferably homing endonucleases as described in detail below.

2. Artificial nucleases as described in detail below, such as, for example, chimeric nucleases, mutated restriction or homing endonucleases or RNA protein particles derived from mobile introns of group II.

Both natural and artificially produced DSBI enzymes are suitable. Preferred are all those DSBI enzymes whose recognition sequence is known and which are either in the form of their Proteins can be obtained (for example by purification) or expressed using their nucleic acid sequence.

Examples of artificial DSBI enzymes are chimeric nucleases 5, which are composed of an unspecific nuclease domain and a sequence-specific DNA binding domain (eg consisting of zinc fingers) (Smith J et al. (2000) Nucl Acids Res 28 (17): 3361 -3369; Bibikova M et al. (2001) Mol Cell Biol. 21: 289-297; Chandrasegaran S and Smith J (1999) Biol Chem 10 380: 841-848; Kim YG and Chandrasegaran S (1994) Proc Natl Acad Sei USA 91: 883-887; Kim YG et al. (1996) Proc Natl Acad Sei USA 93: 1156-1160; Nahon E and Raveh D (1998) Nucl Acids Res 26: 1233-1239).

The DSBI enzyme is preferably selected with knowledge of its specific recognition sequence in such a way that, in addition to the target recognition sequence, it has no other functional recognition regions in the piastome. Homing endonucleases are therefore particularly preferred (overview: Beifort M and Roberts RJ (1997) Nucleic

20 Acids Res 25: 3379-3388; Jasin M (1996) Trends Genet 12: 224-228; Internet: http: //rebase.neb.com/rebase/rebase.homing.html; Roberts RJ and Macelis D (2001) Nucleic Acids Res 29: 268-269). Due to their long recognition sequences, these meet this requirement. However, due to the small size of the piastome, it is also conceivable that DSBI enzymes with shorter recognition sequences (for example restriction endonucleases) can be used successfully.

0 are very particularly preferred

I-NanI (Eide M et al. (1999) Eur J Biochem. 259: 281-288; GenBank Acc.-No .: X78280, nucleotides 418 to 1155),

I-NitI (GenBank Acc.-No .: XI 8211, nucleotides 426 to 1163), 5 I-Njal (GenBank Acc.-No .: X78279, nucleotides 416 to 1153),

I-Ppol encodes the extrachromosomal DNA in the core of Physarum polycephalu (Muscarella DE and Vogt VM (1989) Cell 0 56: 443-454; Lin J and Vogt VM (1998) Mol Cell Biol

18: 5809-5817; GenBank Acc.-No .: M38131, nucleotides 86 to 577; see also SEQ ID NO: 4),

I-Pspl (GenBank Acc.-No .: U00707, nucleotides 1839 to 3449), 5 I-Scal (Monteilhet C et al. (2000) Nucleic Acids Res 28: 1245-1251; GenBank Acc.-No .: X95974, nucleotides 55 to 465)

Endo Scel (Kawasaki et al. (1991) J Biol Chem 266: 5342-5347, identical to F-Scel; GenBank Acc.-No .: M63839, nucleotides 159 to 1589),

I-SceII (Sarguiel B et al. (1990) Nucleic Acids Res 18: 5659-5665),

I-SceIII (Sarguiel B et al. (1991) Mol Gen Genet. 255: 340-341),

I-Ssp6803l (GenBank Acc.-No .: D64003, nucleotides 35372 to 35824),

I-Tevl (Chu et al. (1990) Proc Natl Acad Sei USA 87: 3574-3578; Bell-Pedersen et al. (1990) Nucleic Acids Resl8: 3763-3770; GenBank Acc. -No.: AF158101, nucleotides 144431 until 143694),

I-TevII (Bell-Pedersen et al. (1990) Nucleic Acids Res 18: 3763-3770, -GenBank Acc.-No .: AF158101, nucleotides 45612 to 44836),

I-TevIII (Eddy et al. (1991) Genes Dev. 5: 1032-1041),

Commercially available homing endonucleases such as I-Ppol, PI-PspI or Pl-Scel are very particularly preferred. I-Ppol is the most preferred.

The homing endonuclease according to the protein sequences described by SEQ ID NO: 5 is most preferred. In the production of corresponding expression cassettes, therefore, nucleic acid sequences are used which code for a protein according to SEQ ID NO: 5, the use of the nucleic acid sequences according to is particularly preferred SEQ ID NO: 4.

Restriction endonucleases (type II) can also be used advantageously as DSBI enzymes if they have no interfaces in the piastome. The restriction endonucleases Notl, CciNI, Fsel, Sbfl, Sdal and Sse8387I are particularly preferred.

DSBI enzymes are particularly preferably expressed in accordance with the methods described above for recombinases in such a way that their function in the plastids is ensured. Expression as a fusion protein with a PLS or integration of an expression cassette into the plastid DNA is particularly preferred.

"Recognition sequence for the targeted induction of DNA double-strand breaks" (as a result of "DSB recognition sequence" for double-strand break recognition sequence) generally means those sequences which, under the conditions in the plastids of the plant cell or plant used in each case, the detection and cleavage by a DSBI enzyme allow. DSB recognition sequences for homing endonucleases which are naturally encoded in mitochondria or nucleus of other organisms are particularly preferred. DSB recognition sequences of the homing endonucleases derived from plastids (for example from green algae) can also be used. Preferably, the DSB recognition sequence is singular in the plastid DNA, i.e. a double strand break is only generated at the predefined point. However, cases are also conceivable in which there is more than one DSB recognition sequence in the piastom. This is particularly the case if the DSB recognition sequence is located in duplicate genes (e.g. in inverted "repeats"). In the latter case, there is more than one identical DSB recognition sequence, but their context is identical. It is even preferred that the DSB recognition sequence is present in all copies of the "repeats" so that a cut can also be made in all copies. DSB recognition sequences that occur more than once in a piastom but are localized in the same plastomic context (e.g. in "repeats" or gene duplications) are subsumed within the scope of this invention under the term singular DSB recognition sequences.

Preferably, the plant used or the cell derived from it is predominantly homoplastomic or homotransplastomic with respect to the DSB recognition sequence, i.e. that the majority of the plastid DNA molecules contained in a plastid have this DSB recognition sequence. Such plants are also referred to as master plants in the context of this invention.

Table 2: Detection sequences and organism of origin of DSBI enzymes (" ^Λ " indicates the interface of the DSBI enzyme within a detection sequence.)

Construction of the transformation construct with the insertion sequence

On the transformation construct, which is used to transform the plastids of said master plants, there is at least one RE sequence R2 which, under the action of the recombinase, can recombine with a RE sequence R1 in the master plant.

In a preferred embodiment, the recombination of R1 with R2 leads to hybrid recombinase recognition sequences (R1 / 2 or R2 / 1) which themselves represent only poor or no recombination substrates for the recombinase alone. Said hybrid sequences can certainly be substrates for the recombinase in question if additional factors are added (e.g. the factor xis and IHF when using the λ recombinase). However, these further factors are preferably not present in the plant organism at the same time as the recombinase. The speed of a recombination reaction using the RE sequences R1 and / or R2 is preferably at least ten times as high as for the use of the hybrid sequences R1 / 2 and / or R2 / 1, preferably at least one hundred times as high, particularly preferably at least one thousand times as high , most preferably at least ten thousand times as high. As a result, the balance of the recombination lies strongly on the insertion side, so that the reaction is quasi irreversible.

In a preferred embodiment, the RE sequence (for example the RE sequence R1, R1 ', R2 or R2') - in the 5 '/ 3' direction - comprises a first sequence R-5 ', followed by a core sequence and a second sequence R-3 '. A particularly preferred form of these sequences comprises the loxP sequence, the FRT sequence or the attB or attP of the ΦC31 or TP901-1 recombinase.

In a further preferred embodiment, at least one of the two RE sequences is a pseudo RE sequence or a hybrid RE sequence.

If, for example, the recombinase is selected from the group of resolvases / invertases such as — particularly preferably — C31 recombinase, TP901-1 recombinase or R4 recombinase, the sequences R1 and R2 are preferably not identical. The sequences R1 and R2 preferably correspond to the sequences attB and attP, or vice versa. Furthermore, at least one of the two sequences attB or attP can be replaced by a corresponding pseudo-attB or pseudo-attP RE sequence. The attB sequence preferably comprises - in the 5 '/ 3' direction - a first DNA sequence (attB-5 '), followed by a core region and a second sequence (attB-3').

The attP sequence preferably comprises - in the 5 '/ 3' direction - a first DNA sequence (attP-5 '), followed by a core region and a second sequence (attP-3').

An embodiment is particularly preferred in which the recombinase-mediated recombination between attB and attP produces a recombination product which can no longer act as a substrate for the recombinase alone (i.e. without the participation of additional factors). This recombination product preferably contains a region consisting of - in the 5 '/ 3' direction - a first DNA sequence (attB-5 '), followed by the recombination result of the two core regions and a second sequence (attP-3') and a second Region consisting of - in the 5 '/ 3' direction - a first DNA sequence (attP-5 '), followed by the recombination result of the two core regions and a second sequence (attB-3').

Particularly preferably, at least one of the two RE sequences attB and / or attP is a pseudo-RE sequence (attφ) comprising - in the 5 '/ 3' direction - a first DNA sequence (attφ-5 '), followed by a core region and a second sequence (attφ-3 '). The sequences resulting from the recombination include, for example, a region consisting of - in the 5 '/ 3' direction - a first DNA sequence (attφ-5 '), followed by the recombination result of the two core regions and a second sequence (attP- 3 ') and a second region consisting of - in the 5' / 3 'direction - a first DNA sequence (attP-5'), followed by the recombination result of the two core regions and a second sequence (attφ-3 '), whereby the resulting regions can no longer act as a substrate for the recombinase, making the insertion virtually irreversible. Alternatively, of course, attB and attφ can also be recombined accordingly.

For example, when using the Cre recombinase, the recombination of the RE sequences lox66 and lox71 (see above) leads to a wild-type loxP RE sequence and the lox72 RE sequence

(5 '-taccgttcgtata gcatacat tatacgaacggta-3'), which can only recombine with one another very inefficiently (Araki K et al. (1997) Nucl Acids Res 25: 868-872). Likewise, the recombination of the RE sequences lox76 and lox75 (see above) leads to a wiltyp loxP RE sequence and the lox72 RE sequence (5 '-taccgggcgtata gcatacat tatacgcccggta-3'), which can only recombine with one another very inefficiently. Likewise, the recombination of the RE sequences lox43 and lox44 for wild-type loxP RE sequence and the lox65 RE sequence (5-aatgcatgctata gcatacat tataggtaccgag-3 '), which can only recombine with one another very inefficiently (Albert H et al. (1995) Plant J 7: 649-659). In general, "right element (RE) mutant lox sites" and "left element (LE) mutant lox sites" should be established for use in the context of this invention, which can recombine with one another. The resulting loxP (wild-type) RE sequence and the RE / LE mutated RE sequence only recombine very poorly. This is mainly due to the fact that the LE / RE mutant can no longer be recognized as a substrate.

In particular, the TP901-1 recombinase, the ΦC31 recombinase, the R4 recombinase are recombinases that could be used for almost irreversible recombination. Also possible are λ integrase (preferably with coexpression of IHF), Cre (with correspondingly mutated RE sequences, see above), Flp (with correspondingly mutated RE sequences) and the L5 recombinase.

The transformation construct can also contain a selection marker. If there is already a selection marker in the master plant, said selection markers preferably differ. In addition, the transformation construct preferably comprises one or more genes of interest.

Regulatory elements can also be present on the transformation construct, the RE construct has already been introduced into the master plant, or endogenous regulatory elements that are naturally present at the integration site can be used.

The transformation construct is preferably contained in a transformation vector, which can comprise further elements, for example for replication in bacteria.

If the transformation construct contains only a single RE sequence R2, the transformation construct preferably represents a circular DNA molecule. In this case, the transformation construct can be identical to the transformation vector. However, it is not necessary that the vector sequences (e.g.

Origin of replication for the multiplication of the DNA in bacteria and possibly. a bacterial selection marker) can also be integrated into the plastid genome. These sequences of the vector backbone are preferably removed before the transformation. For example, the insertion sequence can be surrounded by two recognition sites for restriction enzymes. After duplicating the corresponding vector in a bacterium such as E. coli, the Then treat isolated DNA with the appropriate restriction endonuclease or the corresponding restriction endonucleases in vitro. This separates the insertion sequence from the vector backbone and can be circularized again in vitro by using a ligase. The isolated and circularized insertion sequence can then be used for the transformation. Alternatively, the insertion sequence can also be amplified and circularized by means of a polymerase chain reaction using suitable primers.

Alternatively and particularly preferably, the vector backbone is removed from the insertion sequence with the aid of recombinases. For this purpose, the fragment to be transformed is surrounded with two recognition sites for a recombinase, with the aid of which the fragment to be transformed can be excised (for example Cre-lox, FLP / frt). It is possible to design the approaches in such a way that the action of the recombinase only produces the RE sequence R2 necessary for the integration of the insertion sequence.

It is also possible to transform the complete transformation vector and integrate it using recombinases. In this case, the vector backbone sequences are then preferably eliminated again by suitable excision methods.

In order to further increase the efficiency of the method of integration according to the invention by means of recognition site-specific recombination, the number of copies of the DNA molecules to be transformed can be increased, for example, by the transformation construct comprising elements (for example a plastid ORI origin of replication, origin of replication), which enable him to replicate autonomously in the plastid before integration into the plastid DNA or to exist stably as an extrachromosomal DNA molecule in the plastids. Corresponding methods are known to the person skilled in the art (US 5,693,507; US 5,932,479; WO 99/10513). This method is preferred because it increases the number of copies of the insertion sequences available for integration in the plastid.

Alternatively, a "replacement" strategy can also be used. For this purpose, two RE sequences Rl and Rl 'which cannot recombine with one another under the action of the corresponding recombinase or recombinases are introduced into the master plants (or sequences which are naturally present there) are introduced. The transformation construct then contains recognition sequences R2 and R2 'which do not recombine with one another under the action of the recombinase or recombinases, but Rl can recombine with R2 and Rl' with R2 '. As a result, the sequence between Rl and Rl ' in the master plant replaced by the sequence between R2 and R2 'on the transformation construct. The "replacement" strategy not only makes it possible to avoid vector backbone insertions, it also enables the insertion sequence (plus the portions of the recognition sequences R2 and R2 ') to be introduced as a linearized fragment. Such a "replacement" strategy has been demonstrated, for example, in mouse cells with the aid of different recognition regions for the Cre recombinase (Bethke B and Sauer B (1997) Nucl Acids Res 25: 2828-2834). The "Replacement" strategy was also at the core of

Schizisacchromyces pombe realized with the recombinase ΦC31 (Thomason LC (2001) Mol Gen Genet 265: 1031-1038). If the "replacement" strategy is used, a selection marker possibly introduced with the RE construct is preferably positioned between the recognition regions Rl and Rl ', so that it is eliminated directly from the master plant when the insertion sequence is flanked by R2 and R2'.

In the "Replacement" strategy, 2 RE sequences (R1 and R1 'or R2 and R2') can preferably be used (for example 2 lox

Sequences) can be used that cannot recombine with each other because they differ, for example, in the spacer

(see also Bethke B et al. (1997) Nucl Acids Res 25: 2828-2834).

The same has also been described for the FLP / frt, in which in particular most of the nucleotides of the spacer region can be exchanged (Schlake, T & Bode, J, 1994, Biochemistry 33:

12746-12751).

2 identical attB or attP sites can also be used for the replacement strategy. These cannot recombine with each other.

In a further — particularly preferred embodiment — the RE construct preferably contains at least one recognition site for a DSBI enzyme in addition to the recognition site for a recombinase. As a result, these constructs are also referred to as RE / DSB constructs, but are also subsumed under the general term of the RE construct.

The respective DSBI enzyme has - apart from the one introduced via the RE / DSB construct - preferably no further DSB recognition sites in the piastom (except: duplications, for example in "repeats"). It is not imperative that the DSB recognition sequence is present on the RE / DSB construct. It can also be naturally present in the piastome adjacent to the insertion site of a RE construct. A functional RE / DSB unit is only created after the inte- the RE construct into the piastom. With the help of the RE / DSB construct, a predominantly homoplastome master plant is first produced, using methods familiar to the person skilled in the art, such as, for example, homologous recombination.

In a further embodiment, the insertion sequence additionally introduces a sequence X into the piastome which is homologous to a sequence X 'of the master plant which is already present in the piastome. This sequence X 'can be naturally present, or it can be introduced by the RE / DSB construct. The sequences X and X 'are localized in such a way that the following elements lie between them - after integration of the insertion sequence by means of recombination:

i) a preferably singular DSB recognition sequence for a DSBI enzyme

ii) hybrid RE sequence Rl / 2 of the recombinase, which resulted after the incorporation of the secondary transformation vector by recombining Rl with R2

iii) if necessary. other elements such as positive or negative selection markers

wherein X and X 'have an orientation and sufficient homology and length to homologously recombine with one another as a result of a sequence-specific double-strand break on the DSB recognition sequence and to cause a deletion of the sequences located between them.

The transformation construct, the transformation vector or the insertion sequence are introduced into the plastids as described above. The insertion sequence is built into the piastome by the activity of a corresponding recombinase. A system is preferably used which is almost or completely irreversible in the form used. The recombinases of the phages ΦC31 or TP901-1 and their corresponding recognition regions are particularly preferably used.

In a particularly preferred embodiment, after integration of the insertion sequence, a corresponding DSBI enzyme is induced, which produces a DNA double-strand break on its DSB recognition sequence in the piastome of the master plant. There are two different cases:

a) The copy of the plastid DNA under consideration has already been modified in the desired manner by insertion of the insertion sequence mediated by recombinase. The DNA double-strand break occurs at a recognition region that is located adjacent to said insertion sequence. Through intramolecular homologous recombination between X and X, the DSB recognition site located between them is now removed or at least destroyed in such a way that it is no longer recognized by the DSBI enzyme. As a result, the modified plastid DNA molecules can no longer be cut by the DSBI enzyme. They can replicate without interference and / or act as substrates for repair synthesis of further - as yet unmodified - plastid DNA molecules. In a variant of this embodiment, the induced deletion deletes a negative selection marker, which enables easier selection of the plastid DNA molecules with deletion.

b) The copy of plastidic DNA under consideration has not yet been modified in the desired manner by recombinase-mediated insertion of the insertion sequence. In this case, too, there is a DNA double strand break. However, no homologous region is available intramolecularly that could be used to repair the double-strand break. No intramolecular recombination between X and X 'is possible, and consequently no deletion of the DSB recognition sequence. The plastidic DNA molecules cut in this way can on the one hand no longer replicate undisturbed. On the other hand, said cut molecules cannot undergo a repair synthesis by means of a gene conversion event using copies of the plastidic DNA molecules which have already been modified in the desired manner and which can no longer be cut (ie as a result of the events described under a) above, no DSB recognition sequence more included) to be repaired. In this way, even without selective regeneration, the transformation construct / insertion sequence can be distributed into all or at least the majority of plasti copies of a plastid. This procedure speeds up the process

Process of plastid transformation and reduces the risk of accumulation of somaclonal variations due to less necessary regeneration steps. It also enables distribution of the transgene in the copies of the piastome to be achieved without having to use a selection marker and corresponding selective agent. The transformation construct can include further elements. It may be desirable to place at least one gene for a positive and / or negative selection marker between sequence region X and the DSB recognition sequence. The induced recombination also eliminates the selection markers or any other sequence section that is located between the homologous regions X and X '. Accordingly, the vector backbone can be eliminated from the piastom again, for example, if it was not removed before the insertion.

The endonuclease I-Ppol from Physum polycephalum is preferably used as the DSBI enzyme. However, it is also possible to use any other DSBI enzyme that is active in the organelle under consideration. The DSBI enzyme is preferably activated only after the recombinase, which is responsible for the incorporation of the transformation vector / the insertion sequence. Here too, various options are conceivable for realizing this. Corresponding methods have already been described for the recombinase and can be transferred to the DSBI enzyme. The DSBI enzyme can be expressed in fusion with a transit peptide in the core of the master plant. The expression should preferably be inducible or, in the case of constitutive expression, the activity of the corresponding endonuclease should be activatable. In addition, the DSBI enzyme can be expressed as a fusion with a transit peptide in agrobacteria and introduced into plant cells as a fusion protein, for example with VirF. Transient expression from the nucleus with post-translational import into the plastids under consideration is also conceivable. However, the DSBI enzyme or a nucleic acid sequence encoding the DSBI enzyme is preferably introduced into the plastids of the master plant only with the transformation vector. This can be done by introducing the DSBI enzyme as a protein into the organelles at the same time as the transformation construct / insertion sequence; that an in vitro generated RNA, which encodes the DSBI enzyme, is introduced into the plastids with the transformation construct / insertion sequence; or that - particularly preferably - a DNA coding for the DSBI enzyme is introduced into the plastids with the transformation construct / insertion sequence. In the latter case, the DSBI enzyme is very particularly preferably encoded in a form which can be expressed in the plastid. The DSBI enzyme is particularly preferably encoded on the insertion sequence. The expression of the DSBI enzyme is very particularly preferably made possible only after the insertion of the insertion sequence. This can be achieved, for example, by the fact that there are no regulatory elements on the transformation construct / insertion sequence which control the expression of the DSBI enzyme. Expression of the DSBI enzyme could be in one In such a case, for example, after integration into the plastid DNA, it can also be controlled by regulatory elements from the piastome of the master plant. If the expression cassette for the DSBI enzyme is contained on the insertion sequence, the DSBI enzyme is - particularly preferably - flanked by areas which permit later excision of the DSBI enzyme.

In a preferred embodiment, the insertion sequence comprises at least one further nucleic acid sequence to be expressed, which preferably generates a desired phenotype in a correspondingly transformed plant. In order to ensure expression (transcription and / or translation), depending on the embodiment and the place of insertion, these are to be provided with regulatory elements. If the insertion is into a transcriptionally active locus, no promoter sequences are required, as described above. In any case, the sequences to be expressed are advantageously provided with ribosome binding sites at a suitable distance upstream of the open reading frame or already have them by nature. However, these regulatory sequences or parts thereof can naturally also be present in the piastome or already in the first step, i.e. when generating a non-natural master plant together with that of the RE sequence are introduced into the plastid DNA.

"Sufficient length" means in relation to the homology sequences (eg A, A ', B, B', X, X ', Hl, H2) preferably sequences of a length of at least 20 base pairs, preferably at least 50 base pairs, particularly preferably of at least 100 base pairs, very particularly preferably of at least 250 base pairs, most preferably of at least 500 base pairs.

"Adequate homology" in relation to the homology sequences (for example A, A ', B, B', X, X ', Hl, H2) preferably means sequences which have a homology within these homology sequences of at least 70%, preferably 80%, preferably at least 90%, particularly preferably at least 95%, very particularly preferably at least 99%, most preferably 100% over a length of at least 20 base pairs, preferably at least 50 base pairs, particularly preferably of at least 100 base pairs particularly preferably of at least 250 base pairs, most preferably of at least 500 base pairs.

Homology between two nucleic acids is understood to mean the identity of the nucleic acid sequence over the respective total sequence length, which can be determined by comparison using the program algorithm GAP (Wisconsin Package Version 10.0, University of Wisconsin, Genetics Computer Group (GCG), Madison, USA) using the following parameters:

Gap Weight: 12 Length Weight: 4

Average Match: 2,912 Average Mismatch: -2, 003

The transformation construct or the insertion sequence preferably comprises a selection marker which enables a selection of trans-plastomer plastids (see below), particularly preferably aadA, BADH or a "binding type" marker. The selection marker is preferably constructed so that a subsequent deletion from the piastome is made possible. Corresponding methods are known to the person skilled in the art and are described below.

The insertion sequence, the transformation construct, the RE construct or the expression cassettes for a recombinase or a DSBI enzyme can be cloned into a standard vector such as pBluescript or pUC18 in order to construct a transformation vector. In a preferred embodiment, the insertion sequence or transformation construct is applied as a linear or linearized DNA molecule.

Preferably only the part of the transformation vector is applied which comprises the insertion sequence or the transformation construct with RE sequence, selection marker and / or the expression cassette for the DSBI enzyme.

One of the constructs described above can be introduced into the plastids of a corresponding master plant using one of the methods described. Microinjection and particle bombardment are preferred.

Cloning, expression, selection and transformation procedures

“Expression cassette” means - for example in relation to the expression cassette for a recombinase or a DSBI enzyme - such constructions in which the DNA to be expressed is functionally linked to at least one genetic control element that its expression (ie transcription and or translation ) enables or regulates. For example, the expression can be stable or transient, constitutive or inducible. Various direct (e.g. transfection, particle bombardment, microinjection) or indirect (e.g. agrobacterial) infection, viral infection) are available below.

A functional link is generally understood to be an arrangement in which a genetic control sequence can perform its function in relation to a nucleic acid sequence - for example coding for a recombinase or a DSBI enzyme. Function can, for example, control expression, i.e. Transcription and / or translation of the nucleic acid sequence - for example coding for a recombinase or a DSBI enzyme - mean. Control includes, for example, the initiation, increase, control or suppression of expression, i.e. Transcription and, if necessary, translation. The control, in turn, can take place in a tissue-specific or time-specific manner, for example. It can also be inducible, for example, by certain chemicals, stress, pathogens, etc.

A functional link is understood to mean, for example, the sequential arrangement of a promoter, the nucleic acid sequence to be expressed - for example coding for a recombinase or a DSBI enzyme - and possibly. further regulatory elements such as a terminator such that each of the regulatory elements can fulfill its function in the expression of the nucleic acid sequence - for example coding for a recombinase or a DSBI enzyme. It is not absolutely necessary that the functional link is already given on the transformation constructs. The functional linkage can also result from the insertion into the core or plastid DNA, the regulatory elements already being present in the core or plastid DNA. In this regard, the regulatory elements can naturally exist or can be introduced in an upstream step - for example when introducing a RE sequence.

This does not necessarily require a direct link in the chemical sense. Genetic control sequences, such as, for example, enhancer sequences, can also perform their function on the target sequence from more distant positions or even from other DNA molecules. Arrangements are preferred in which the nucleic acid sequence to be expressed - for example coding for a recombinase or a DSBI enzyme - is positioned behind a sequence which acts as a promoter, so that both sequences are covalently linked to one another. The distance between the promoter sequence and the nucleic acid sequence - for example coding for a recombinase or a DSBI enzyme - is preferably less than 200 base pairs, particularly preferred less than 100 base pairs, very particularly preferably less than 50 base pairs.

Various ways are known to the person skilled in the art in order to arrive at one of the transformation constructs according to the invention, these vectors or one of the expression cassettes. The production can be carried out using common recombination and cloning techniques, as described, for example, in Maniatis T, Fritsch EF and Sambrook J, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1989) and in Silhavy TJ, Berman ML and Enquist LW, Experiments with Gene Fusions, Cold Spring Harbor Laboratory, Cold Spring Harbor, NY (1984) and in Ausubel FM et al. , Current Protocols in Molecular Biology, Greene Publishing Assoc. and Wiley Interscience (1987). The direct fusion of a nucleic acid sequence functioning as a promoter with a nucleotide sequence to be expressed is preferred - for example coding for a recombinase or a DSBI enzyme.

The term "genetic control sequences" is to be understood broadly and means all those sequences which have an influence on the formation or the function of an expression cassette or transformation vector. Genetic control sequences ensure transcription and, if necessary, translation in the cell nucleus (or cytoplasm) or plastids. The expression cassettes according to the invention preferably comprise a promoter 5 'upstream of the respective nucleic acid sequence to be expressed and a terminator sequence 3' downstream as an additional genetic control sequence, and optionally further customary regulatory elements, in each case functionally linked to the nucleic acid sequence to be expressed.

Genetic control sequences are described, for example, in "Goeddel; Gene Expression Technology: Methods in Enzymology 185, Academic Press, San Diego, CA (1990)" or "Gruber and Crosby, in: Methods in Plant Molecular Biology and Biotechnology, CRC Press, Boca Raton , Florida, eds .: Glick and Thompson, Chapter 7, 89-108 "and the quotations cited there.

Examples of such control sequences are sequences to which inducers or repressors bind and thus regulate the expression of the nucleic acid. In addition to these new control sequences or instead of these sequences, the natural regulation of these sequences may still be present before the actual structural genes and may have been genetically modified so that the natural regulation is switched off and the Expression of the genes was increased. However, the expression cassette can also have a simpler structure, ie no additional regulation signals are inserted in front of the genes mentioned above and the natural promoter with its regulation is not removed. Instead, the natural control sequence is mutated so that regulation no longer takes place and gene expression is increased. These modified promoters can also be placed in front of the natural genes to increase activity.

Depending on the plant organisms described in more detail above, which can be used as host or starting organism and which is converted into a genetically modified or transgenic organism by introducing the expression cassettes or vectors, various control sequences are suitable.

In principle, all promoters which can control the expression of genes, in particular foreign genes, in plants are suitable for nuclear expression (for example a viral / bacteriophage RNA polymerase, a recombinase or a DSBI enzyme in each case as a fusion protein with a plastid transit peptide).

Promoters which allow constitutive expression in plants are suitable (Benfey et al. (1989) EMBO J 8: 2195-2202). In particular, a vegetable one is preferably used

Promoter or a promoter derived from a plant virus. The promoter of the 35S transcript of the cauliflower mosaic virus (Franck et al. (1980) Cell 21: 285-294; Odell et al. (1985) Nature 313: 810-812; Shewmaker et al. (1985) Virology 140: 281-288; Gardner et al. 1986, Plant Mol. Biol. 6, 221-228) or the 19S CaMV promoter (US 5,352,605 and WO 84/02913). As is known, this promoter contains different recognition sequences for transcriptional effectors, which in their entirety lead to a largely permanent and constitutive expression of the introduced gene (Benfey et al.

(1989) EMBO J 8: 2195-2202). Another suitable constitutive promoter is the "Rubisco small subunit (SSU)" promoter (US 4,962,028). Another example of a suitable promoter is the LeguminB promoter (GenBank Acc.-No .: X03677). Further preferred constitutive promoters are, for example, the promoter of nopaline synthase from Agrobacterium, the TR double promoter, the OCS (octopine synthase) promoter from Agrobacterium, the ubiquitin promoter (Holtorf S et al. (1995) Plant Mol Biol 29: 637- 649), the promoters of the vacuolar ATPase subunits, the FBPaseP promoter (WO 98/18940) or the promoter of a proline-rich protein from wheat (WO 91/13991). Also suitable and preferred within the scope of this invention are constitutive promoters the SuperPromotor (Ni M et al. (1995) Plant J 7: 661-676; US 5,955,646) and the nitrilase-1 promoter of the nitl gene from Arabidopsis (GenBank Acc. -No.: Y07648.2, nucleotides 2456 to 4340; Hillebrand H et al. (1998) Plant Mol Biol 36 (l): 89-99; Hillebrand H et al. (1996) Gene 170 (2): 197-200).

Preferred are inducible, particularly preferably chemically inducible promoters (Aoyama T and Chua NH (1997) Plant J 11: 605-612; Caddick MX et al. (1998) Nat. Biotechnol 16: 177-180; Rewiew: Gatz (1997) Annu Rev Plant Physiol Plant Mol Biol

48: 89-108) by which expression can be controlled at a particular point in time. Examples include the PRP1 promoter (Ward et al. (1993) Plant Mol Biol 22: 361-366), a promoter induced by salicylic acid (WO 95/19443), a promoter induced by benzenesulfonamide (EP-A-0388186) , a tetracycline-inducible promoter (Gatz et al. (1992) Plant J 2: 397-404), an abscisic acid-inducible promoter (EP-A 335 528), an ethanol-inducible promoter (Salter MG et al. ( 1998) Plant J. 16: 127-132), the heavy metal-inducible metallothionein I promoter (Amini S et al. (1986) Mol Cell Biol 6: 2305-2316), the steroid-inducible MMTV LTR promoter (Izant JG et al (1985) Science 229: 345-352) and a cyclohexanone-inducible promoter (WO 93/21334). The inducible expression of a PLS / DSBI enzyme fusion protein in the core is particularly preferred. Inducible promoters also include those that can be regulated by certain repressor proteins (e.g. tet, lac). Corresponding repressor proteins can be translocated into the plastids in fusion with PLS and there regulate the expression of certain genes under the control of appropriate promoters. The repressor binds in the plastids to an artificial repressor binding site inserted into the piastome and can thus suppress the expression of the gene located downstream (cf. W095 / 25787). In this way, for example, the expression of a plastid-coded recombinase or a DSBI enzyme can be induced if necessary or suppressed until the moment when expression is desired.

Furthermore, promoters are preferred which are induced by biotic or abiotic stress, such as, for example, the pathogen-inducible promoter of the PRPl gene (Ward et al.,

Plant Mol Biol 1993, 22: 361-366), the heat-inducible hsp70 promoter or the hsp80 promoter from tomato (US 5,187,267), the cold-inducible alpha-amylase promoter from the potato (WO 96/12814) or the wound-induced pinll promoter (EP-A 0 375 091). In a particularly preferred embodiment, especially the nucleic acid coding for the recombinase or the DSBI enzyme is expressed under the control of an inducible promoter. This achieves controlled, controllable expression and avoids any problems caused by constitutive expression of a recombinase or a DSBI enzyme.

Advantageous control sequences for the expression cassettes or vectors according to the invention include viral, bacteriophage or bacterial promoters such as cos, tac, trp, tet, phoA, tat, lpp, lac, laclq, T7, T5, T3 -, gal, trc, ara, SP6, λ-PR or in the λ-PL promoter. These are preferably used in combination with the ^'expression of the respective corresponding RNA polymerase.

Expression in plastids can be realized using plastid promoters and / or transcription regulatory elements. The RNA polymerase promoter (WO 97/06250) or those in WO 00/07431, US 5,877,402, WO 97/06250, WO 98/5559,5 WO 99/46394, WO 01/42441 and Promoters described in WO 01/07590. These include the rpo B promoter element, the atpB promoter element, the clpP promoter element (see also WO 99/46394) or the 16S rDNA promoter element. A polycistronic "operon" can also be assigned to the promoter (EP-A 1 076 095; WO 00/20611). Systems are also described in which a non-vegetable (eg viral) RNA polymerase is imported into the plastid using plastid transit peptides and specifically induces the expression of transgenic sequences there, which are under the control of the recognition sequences of the RNA polymerase and previously into the plastid one DNA were inserted (WO 95/16783; US 5,925,806; US 5,575,198).

In addition to the promoters mentioned, it would also be preferable to use:

a) the PrbcL promoter (SEQ ID NO: 20)

b) the Prpsl6 promoter (SEQ ID NO: 26)

c) the Prrnl6 promoter (SEQ ID NO: 22)

In a preferred embodiment, NEP promoters are used. These are promoters that are functional in plastids and are recognized by the nuclear-encoded plastid RNA polymerase (NEP). Preferred are: Prrn-62; Pycf2-1577; PatpB-289; Prps2-152; Prpsl6-107; Pycfl-41; PatpI-207; PclpP-511; PclpP-173 and PaccD-129 (WO 97/06250; Hajdukiewicz PTJ et al (1997) EMBO J 16: 4041-4048).

The following are particularly preferred:

a) The PaccD-129 promoter of the accD gene from tobacco (WO 97/06250; SEQ ID NO: 23)

b) The PclpP-53 promoter of the clpP gene as a highly active NEP promoter in chloroplasts (WO 97/06250; SEQ ID NO: 24)

c) The Prrn-62 promoter of the rrn gene (SEQ ID NO: 25)

d) The Prpsl6-107 promoter of the rpslδ gene (SEQ ID NO: 21):

e) The PatpB / E-290 promoter of the atpB / E gene from tobacco (Kapoor S et al. (1997) Plant J 11: 327-337) (SEQ ID NO: 27)

f) The PrpoB-345 promoter of the rpoB gene (Liere K & Maliga P (1999) EMBO J 18: 249-257) (SEQ ID NO: 28)

In general, all promoters belonging to class III can be used in this embodiment (Hajdukiewicz PTJ et al (1997) EMBO J 16: 4041-4048) as well as all fragments of the class II promoters which control the transcription initiation by the NEP. Such promoters or promoter parts are not particularly highly conserved. As consensus near the transcription initiation site of NEP promoters is given: ATAGAATAAA (Hajdukiewicz PTJ et al (1997) EMBO J 16: 4041-4048).

Normally, genes are surrounded by regulatory sequences that come from the plastids of the organism to be transformed. This creates sequence duplications that can lead to instabilities due to spontaneous, intrachromosomal homologous recombination events (Heifetz PB (2000)

Biochimie 82 (6-7): 655-666). To solve this problem, it has been proposed to use heterologous regulatory sequences or to use regulatory units already present endogenously in the plastid genome (WO 99/46394; WO 01/42441). A reduction in homology by mutagenesis of the endogenous promoter sequence is also described (WO 01/07590).

In principle, all natural promoters with their regulatory sequences such as those mentioned above can be used for the method according to the invention. Promoters isolated from prokaryotes are particularly preferred. Promoters isolated from Synechocystis or E. coli are very particularly preferred. About that In addition, synthetic promoters can also be used advantageously, such as a synthetic promoter derived from the E. coli consensus sequence for σ70 promoters:

5-TTGACA N ₁₆ _ ₁₉ TATAAT N ₃ CAT -3 ',

where N stands for any nucleotide (ie A, G, C or T). It is obvious to the person skilled in the art that individual to few base exchanges are also possible in the specified, conserved regions without destroying the function of the promoter. The variable design of these synthetic promoters by using different sequence sequences makes it possible to create a large number of promoters which do not have extensive homologies, which in particular in the case that several promoters are required, the stability of the expression cassettes in the

Piastom increased. The following particularly preferred promoter sequences, which are derived from the above-mentioned consensus sequence, may be mentioned by way of example but not by way of limitation:

a) 5'-TTGACATTCACTCTTCAATTATCTATAATGATACA-3 '(SEQ ID NO: 29) b) 5'-TTGACAATTTTCCTCTGAATTATATAATTAACAT-3' (SEQ ID NO: 59)

It is obvious to the person skilled in the art that these synthetic promoters can control the expression of any genes. For example, they can be used to drive the expression of a selection marker - also in order to be able to select transplastomic plants under regenerative conditions with the aid of the said selection system. Selection markers are listed below as examples. In addition, such synthetic promoters can be linked to any gene, for example with genes coding for antibodies, antigens or enzymes. The expression cassettes consisting of such promoters preferably also contain 5 'untranslated regions (or ribosome binding sites) or 3' non-coding regions described in more detail below.

Genetic control sequences also include further promoters, promoter elements or minimal promoters that can modify the expression-controlling properties. Genetic control sequences also include the 5 'untranslated region (5'-UTR) or the non-coding 3' region (3'-UTR) of genes (Eibl C (1999) Plant J 19: 1-13). It has been shown that these can play a significant role in regulating gene expression in plastids of higher plants. At its core, genetic control elements such as 5'-UTR, introns or 3'UTR can also have a function in gene expression. For example, it was shown that 5 'untranslated sequences can enhance iente expression of heterologous genes. They can also promote tissue specificity (Rouster J et al. (1998) Plant J 15: 435-440).

The following are preferably used as 5 'UTRs and 3'UTRs:

a) 5'psbA (from tobacco) (SEQ ID NO: 30)

b) 5'rbcL including 5 'portions from the coding region of the rbcL gene (from tobacco) (SEQ ID NO: 31); the sequence described with SEQ ID NO: 31 was mutated compared to the native sequence in order to introduce a Pagl or Ncol interface.

c) 5'rbcLs (SEQ ID NO: 32); that described with SEQ ID NO: 32

Sequences were mutated against the native sequence to introduce a Pagl site.

d) 3'psbA-1 from Synechocystis (SEQ ID NO: 33)

e) 3'psbA from tobacco (SEQ ID NO: 34)

f) 3'rbcL from tobacco (SEQ ID NO: 35)

Genetic control sequences, especially for expression in plastids, in particular also include ribosome formation sequences for the initiation of translation. These are usually included in the 5 'UTRs. This is particularly preferred if the nucleic acid sequence to be expressed does not provide appropriate sequences or if these are provided with the

Expression system are not compatible. Particularly preferred is the use of a synthetic Ribosomenbinde- site (RBS) having the sequence 5'-GGAGG (N) ₃ _ ₁₀ ATG-3 ', preferably, 5'-GGAGG (N) ₅ ATG-3' (SEQ ID NO: 36 ) particularly preferably 5'-GGAG-GATCTCATG-3 '(SEQ ID NO: 37).

The expression cassette can advantageously contain one or more so-called "enhancer sequences" functionally linked to the promoter, which enable increased transgenic expression of the nucleic acid sequence. Additional advantageous sequences, such as further regulatory elements or terminators, can also be inserted at the 3 'end of the nucleic acid sequences to be expressed transgenically. The nucleic acid sequences to be expressed transgenically can be contained in one or more copies in the gene construct. It is also possible to insert a so-called downstream box after the start codon, which generally increases expression (translation enhancer "WO 00/07431; WO 01/21782).

As genetic control sequences especially Polyadenylation signals suitable for core transformation are plant polyadenylation signals, preferably those which essentially correspond to T-DNA polyadenylation signals from Agrobacterium tumefaciens, in particular gene 3 of T-DNA (octopine synthase) of the Ti plasmid pTiACHS (Gielen et al., EMBO J. 3 (1984), 835 ff) or functional equivalents thereof. Examples of particularly suitable terminator sequences are the OCS (octopine synthase) terminator and the NOS (nopalin synthase) terminator.

The transformation vectors and insertion sequences according to the invention can comprise further nucleic acid sequences. Such nucleic acid sequences can preferably represent expression cassettes. The following may be mentioned as examples, but not restrictive, for the DNA sequences to be expressed in the expression constructs:

1. Selection marker

“Selection marker” means all those nucleic acid or protein sequences whose expression (ie transcription and possibly translation) gives a cell, tissue or organism a different phenotype than an untransformed one. Selection markers include, for example, those nucleic acid or protein sequences whose expression gives a cell, tissue or organism an advantage (positive selection marker) or disadvantage (negative selection marker) over cells which do not express this nucleic acid or protein. Positive selection markers work, for example, by detoxifying a substance that has an inhibitory effect on the cell (e.g. antibiotic / herbicide resistance), or by forming a substance that enables the plant to improve regeneration or increase growth under the selected conditions (for example nutritive markers, hormone-producing markers such as ipt; see below). Another form of positive selection marker includes mutated proteins or RNAs that are not sensitive to a selective agent (for example, 16S rRNA mutants that are insensitive to spectinomycin). Negative selection markers work, for example, by catalyzing the formation of a toxic substance in the transformed cells (for example the codA gene). Furthermore, selection markers can also include reporter proteins, insofar as these are suitable, transformed from non- to differentiate transformed cells, tissues or organs (for example by coloring or another detectable phenotype).

The following selection markers are exemplary and not restrictive:

1.1 Positive selection markers:

The selectable marker introduced into the cell nucleus or the plastids with the expression cassette gives the successfully transformed cells resistance to a biocide (for example a herbicide such as phosphinothricin, glyphosate or bromoxynil), a metabolism inhibitor such as 2-deoxyglucose-6-phosphate (WO 98/45456) or an antibiotic, such as, for example, tetracycline, ampicillin, kanamycin, G 418, neomycin, bleomycin or hygromycin. The selection marker allows the selection of the transformed cells from untransformed ones (McCormick et al., Plant Cell Reports 5 (1986), 81-84; Dix PJ & Kavanagh TA (1995) Euphytica 85: 29-34).

Particularly preferred selection markers are those which confer resistance to herbicides. Examples of selection markers are:

DNA sequences which code for phosphinothricin acetyltransferases (PAT), which acetylate the free amino group of the glutamine synthase inhibitor phosphinothricin (PPT) and thus achieve a detoxification of the PPT (de Block et al. 1987, EMBO J. 6, 2513-2518) (also Bialophos ^® resistance gene (called bar). The bar gene coding for a phosphinothricin acetyltransferase (PAT) can be isolated from, for example, Streptomyces hygroscopicus or S. viridochromogenes. Corresponding sequences are known to the person skilled in the art (from Streptomyces hygroscopicus GenBank Acc.-No .: X17220 and X05822, from Streptomyces viridochromogenes GenBank Acc.-No .: M 22827 and X65195; US 5,489,520). Furthermore, synthetic genes are described, for example, for expression in plastids. A synthetic Pat gene is described in Becker et al. (1994) The Plant J. 5: 299-307. The genes confer resistance to the herbicide bialaphos or glufosinate and are widely used markers in transgenic plants (Vickers JE et al. (1996). Plant Mol Miol Reporter 14: 363-368; Thompson CJ et al. (1987) EMBO J 6: 2519 -2523). 5-enolpyruvylshikimate-3-phosphate synthase genes (EPSP synthase genes) which confer resistance to glyphosate (N- (phosphonomethyDglycine). The unselective herbicide glyphosate has 5-enolpyruvyl-3-phosphoshikimate synthase (EPSPS) as its molecular target has a key role in the biosynthesis of aromatic amino acids in microbes and plants, but not in mammals (Steinrucken HC et al. (1980) Biochem. Biophys. Res. Commun. 94: 1207-1212; Levin JG and. Sprinson DB (1964) J Biol. Chem. 239: 1142-1150; Cole DJ (1985) Mode of action of glyphosate a literature analysis, p. 48-74. In: Grossbard E and Atkinson D (eds.). The herbicide glyphosate. Buttersworths, Boston Glyphosate-tolerant EPSPS variants are preferably used as selection markers (Padgette SR et al. (1996). New weed control opportunities: development of soybeans with a Roundup Ready ™ gene. In: Herbicide Resistant Crops (Duke, SO, ed. ), pp. 53-84.CRC Press, Boca Raton, FL; Saroha MK and Mal ik VS (1998) J Plant Biochemistry and Biotechnology 7: 65-72). The EPSPS gene of the Agrobacterium sp. strain CP4 has a natural tolerance to glyphosate, which can be transferred to corresponding transgenic plants. The CP4 EPSPS gene was derived from Agrobacterium sp. strain CP4 cloned (Padgette SR et al. (1995) Crop Science 35 (5): 1451-1461). Sequences of 5-enolpyrvylshikimate-3-phosphate synthases that are glyphosate tolerant, as described, for example, in US 5,510,471; US 5,776,760; US 5,864,425; US 5,633,435; US 5,627; 061; US 5,463,175; EP 0 218 571 are both described in the patents and deposited in the GenBank. Further sequences are described under GenBank Accession X63374. The aroA gene is also preferred (M10947 S. typhimurium aroA locus 5-enolpyruvylshikimate-3-phosphate synthase (aroA protein) gene).

the gox gene (glyphosate oxidoreductase) coding for the glyphosate ^® degrading enzyme. GOX (e.g. the

Glyphosate oxidoreductase from Achromobacter sp.) Catalyzes the cleavage of a C-N bond in the glyphosate, which is thus converted to aminomethylphosphonic acid (AMPA) and glyoxylate. GOX can thereby confer resistance to glyphosate (Padgette SR et al. (1996) J Nutr. 1996 Mar; 126 (3): 702-16; Shah D et al. (1986) Science 233: 478-481).

the deh gene (coding for a dehalogenase that inactivates Dalapon ^® ), (GenBank Acc. -No.: AX022822, AX022820 and W099 / 27116) bxn genes which code for bromoxynil-degrading nitrilase enzymes. For example, the nitrilase from Klebsiella ozanenae. Sequences can be found in the Genbank, for example, under Acc.-No: E01313 (DNA encoding bromoxynil-specific nitrilase) and J03196 (K. pneumoniae bromoxynil-specific nitrilase (bxn) gene, complete cds).

Neomycin phosphotransferases confer resistance to antibiotics (aminoglycosides) such as neomycin, G418, hygromycin, paromomycin or kanamycin by reducing their inhibitory effect through a phosphorylation reaction. The nptll gene is particularly preferred. Sequences can be obtained from GenBank (AF080390 mini transposon mTn5-GNm; AF080389 mini transposon mTn5-Nm, complete sequence). In addition, the gene is already part of numerous expression vectors and can be isolated from them using methods familiar to the person skilled in the art (such as, for example, polymerase chain reaction) (AF234316 pCAMBIA-2301; AF234315 pCAMBIA-2300, AF234314 pCAMBIA-2201). The NPTII gene encodes an aminoglycoside 3 'O-phosphotransferase from E. coli, Tn5 (GenBank Acc.-No: U00004 position 1401-2300; Beck et al. (1982) Gene 19 327-336) the aphA-6 gene from Acinetoacter baumannii, coding for an aminoglycoside phosphotransferase, can be used as a selection marker (Huang et al. (2002) Mol Genet Genomics 268: 19-27).

the DOG ^ l gene. The gene D0G ^R 1 was isolated from the yeast Saccharomyces cerevisiae (EP 0 807 836). It encodes a 2-deoxyglucose-6-phosphate phosphatase that confers resistance to 2-DOG (Randez-Gil et al. 1995, Yeast 11, 1233-1240; Sanz et al. (1994) Yeast 10: 1195-1202, Sequence: GenBank Acc.-No .: NC001140 Chromosome VIII, Saccharomyces cervisiae Position 194799-194056).

Sulfonylurea and imidazolinone inactivating acetolactate synthases which confer resistance to imidazolinone / sulfonylurea herbicides. The active ingredients imazamethabenz-methyl, imazamox, imazapyr, imazaquin, imazethapyr may be mentioned as examples of imidazolinon herbicides. Examples of sulfonyl urea herbicides are amidosulforon, azim-sulfuron, chlorimuronethyl, chlorosulfuron, cinosulfuron, imazosulforon, oxasulforon, prosulforon, rimsulforon, sulfosulforon. Numerous other active substances of the classes mentioned are known to the skilled worker. Are suitable

Nucleic acid sequences such as, for example, the sequence for Arabidopsis filed under GenBank Acc-No .: X51514 thaliana Csr 1.2 gene (EC 4.1.3.18) (Sathasivan K et al. (1990) Nucleic Acids Res. 18 (8): 2188). Acetolactate synthases that confer resistance to imidazolinone herbicides are also described under GenBank Acc.-No .:

a) AB049823 Oryza sativa ALS mRNA for acetolactate synthase, complete cds, herbicide resistant biotype

b) AF094326 Bassia scoparia herbicide resistant acetolactate synthase precursor (ALS) gene, complete cds

c) X07645 Tobacco acetolactate synthase gene, ALS SuRB (EC 4.1.3.18)

d) X07644 Tobacco acetolactate synthase gene, ALS SuRA (EC 4.1.3.18)

e) A19547 Synthetic nucleotide mutant acetolactate synthase

f) A19546 synthetic nucleotide mutant acetolactate synthase

g) A19545 synthetic nucleotide mutant acetolactate synthase

h) 105376 Sequence 5 from Patent EP 0257993

i) 105373 Sequence 2 from Patent EP 0257993

j) AL133315

- Hygromycin phosphotransferases (X74325 P. pseudomallei gene for hygromycin phosphotransferase) which confer resistance to the antibiotic hygromycin. The gene is part of numerous expression vectors and can be isolated from them using methods familiar to the person skilled in the art (such as, for example, polymerase chain reaction) (AF294981 pINDEX4; AF234301 pCAMBIA-1380; AF234300 pCAM- BIA-1304; AF234299 pCAMBIA-1303; AF234298 p 1302; AF354046 pCAMBIA-1305.; AF354045 pCAMBIA-1305.1)

- resistance genes against

a) chloramphenicol (chloramphenicol acetyl transferase),

b) Tetracycline, various resistance genes are described, e.g. X65876 S. ordonez genes class D tetA and tetR for tetracycline resistance and repressor proteins X51366 Bacillus cereus plasmid pBC16 tetracycline resistance gene. In addition, the gene is already part of numerous expression vectors and can be isolated from these using methods familiar to the person skilled in the art (such as, for example, polymerase chain reaction)

c) Streptomycin, various resistance genes are described e.g. with the GenBank Acc.-No. : AJ278607 Corynebacterium acetoacidophilum ant gene for streptomycin adenylyl transferase.

d) Zeocin, the corresponding resistance gene is part of numerous cloning vectors (e.g. L36849 cloning vector pZEO) and can be isolated from these using methods familiar to the person skilled in the art (such as, for example, polymerase chain reaction).

e) Ampicillin (β-lactamase gene; Datta N, Richmond MH. (1966) Biochem J. 98 (l): 204-9; Heffron F et al (1975) J. Bacteriol 122: 250-256; the Amp gene was first cloned to produce the E. coli vector pBR322; Bolivar

F et al. (1977) Gene 2: 95-114). The sequence is part of numerous cloning vectors and can be isolated therefrom using methods familiar to those skilled in the art (such as, for example, polymerase chain reaction).

- Genes such as isopentenyl transferase from Agrobacterium tumefaciens (strain: P022) (Genbank Acc.-No.: AB025109). The ipt gene is a key enzyme in cytokinin biosynthesis. Its overexpression facilitates the regeneration of plants (e.g. selection on cytokinin-free medium). The procedure for using the ipt gene is described (Ebinuma H et al. (2000) Proc Natl Acad Sei USA 94: 2117-2121; Ebinuma, H et al. (2000) Selection of Marker-free transgenic plants using the oncogenes (ipt , rol A, B, C) of Agrobacterium as selectable markers. In Molecular Biology of Woody Plants. Kluwer Academic Publishers).

Various other positive selection markers which give the transformed plants a growth advantage over untransformed ones, and methods for their use are described, inter alia, in EP-A 0 601 092. Examples include β-glucuronidase (in conjunction with, for example, cytokininglucuronide ), Mannose-6-phosphate-isomerase (in connection with mannose), UDP-galactose-4-epimerase (in connection with eg galactose), whereby Mannose-6-phosphate isomerase is particularly preferred in connection with mannose.

For a selection marker which is functional in plastids, those which are resistant to spectinomycin, streptomycin, kanamycin, lincomycin, gentamycin, hygromycin, methotrexate, bleomycin, phleomycin, blasticidin, sulfonamide,, phosphinotricin, chlorosulfuron, bromoxymil, are particularly preferred. Add glyphosate, 2,4-datrazine, 4-methyltryptophan, nitrate, S-aminoethyl-L-cysteine, lysine / threonine, aminoethyl-cysteine or betaine aldehyde. The genes aadA, nptll, BADH, FLARE-S (a fusion of aadA and GFP, described by Khan MS & Maliga P, 1999 Nature Biotech 17: 910-915) are particularly preferred.

The aadA gene is primarily described as a selection marker functional in plastids (Svab Z and Maliga P (1993) Proc Natl Acad Sei USA 90: 913-917). Modified 16S rDNA, the nptll gene (kanamycin resistance) and the bar gene (phosphinothricin resistance) are also described. Due to the preference of the selection marker aadA, this is preferably "recycled" i.e. after his

Use deleted from the genome or piastome (Fischer N et al. (1996) Mol Gen Genet 251: 373-380; Corneille S et al. (2001) Plant J 27: 171-178), so that aadA is used in further transformations traveled transplastomic plant can be used again as a selection marker. Another possible selection marker is betaine aldehyde dehydrogenase (BADH) from spinach (Daniell H et al. (2001) Trends Plant Science 6: 237-239; Daniell H et al. (2001) Curr Genet 39: 109-116 ; WO 01/64023; WO 01/64024; WO 01/64850). Lethal-acting agents such as glyphosate can also be used in conjunction with corresponding detoxifying or resistant enzymes (WO 01/81605).

Binding type markers can also be used. For example, point mutations can be introduced at a site of the 23SrDNA (position 2073 or 2074 of the 23SrRNA from tobacco, sequence: AAAGACCCTATGAAG) (for example sequence: GG.AGACCCTATGAAG), which confer resistance to lincomycin derived from a mutated 23SrDNA (Cseplö A et al. (1988) Mol Gen Genet 214: 295-299). Other point mutations include those in the 16S rRNA of tobacco that confer resistance to spectinomycin (mutation underlined):

a) 5 '-GGAAGGTGAGGATGC.-3' ("native" here: A)

Other mutations in the 16S rRNA confer resistance to streptomycin: b) 5 '-GAATGAAACTA.-3'("native" here: C)

1.2 Negative selection markers

Negative selection markers allow, for example, the

Selection of organisms with successfully deleted sequences that comprise the marker gene (Koprek T et al. (1999) The Plant Journal 19 (6): 719-726). For example, sequences that code for recombinases, selection markers or for DSBI enzymes can be deleted from the genome / piastome again after successful use of the method according to the invention.

In the case of negative selection, for example, the negative selection marker introduced into the plant converts a compound which otherwise has no adverse effect on the plant into a compound with an adverse effect. Also suitable are genes which have an adverse effect per se, such as, for example, TK thymidine kinase (TK) and Diphtheria Toxin A fragment (DT-A), the codA gene product coding for a cytosine deaminase (Gleave AP et al. (1999) Plant Mol Biol 40 (2): 223-35; Perera RJ et al. (1993) Plant Mol Biol 23 (4): 793-799; Stougaard J (1993) Plant J 3: 755-761), the cytochrome P450 gene ( Koprek et al. (1999) Plant J. 16: 719-726), genes coding for a haloalkane dehalogenase (Naested H (1999) Plant J. 18: 571-576), the iaaH gene (Sundaresan V et al. (1995) Genes & Development 9: 1797-1810) or the tms2 gene (Fedoroff NV & Smith DL (1993) Plant J 3: 273-289).

The concentrations of antibiotics, herbicides, biocides or toxins used for the selection must be adapted to the respective test conditions or organisms.

Examples of plants to be mentioned are kanamycin (Km) 50 mg / L, hygromycin B 40 mg / L, phosphinothricin (Ppt) 6 mg / L, septinomycin (Spec) 500 mg / L.

Furthermore, functional analogs of the nucleic acids mentioned can be expressed coding for selection markers. Functional analogs here means all the sequences that have essentially the same function i.e. are able to select transformed organisms. The functional analogue can differ in other characteristics. For example, it may have a higher or lower activity, or it may have other functionalities.

Functional analogs furthermore mean sequences which code for fusion proteins consisting of one of the preferred selection markers and another protein, for example another preferred selection marker, one of the reporter mentioned below. proteins or a PLS. A fusion of the GFP (green fluorescent protein) and the aadA gene may be mentioned as an example (Sidorov VA et al. (1999) Plant J 19: 209-216).

2. Reporter genes

Reporter genes code for easily quantifiable proteins, which use their own color or enzyme activity to ensure an assessment of the transformation efficiency, the expression site or time or the identification of transgenic plants. Genes coding for reporter proteins are very particularly preferred (see also Schenborn E, Groskreutz D. Mol Biotechnol. 1999; 13 (1): 29-44) such as

"Green fluorescence protein" (GFP) (Chui WL et al., Curr Biol 1996, 6: 325-330; Leffel SM et al., Biotechniques. 23 (5): 912-8, 1997; Sheen et al. ( 1995) Plant Journal 8 (5) -777-784; Haseloff et al. (1997) Proc Natl Acad Sei USA 94 (6): 2122-2127; Reichel et al. (1996) Proc Natl Acad Sei USA 93 (12 ): 5888-5893; Tian et al. (1997) Plant Cell Rep 16: 267-271; WO 97/41228).

chloramphenicol,

- Luciferase (Millar et al., Plant Mol Biol Rep 1992

10: 324-414; Ow et al. (1986) Science, 234: 856-859); allows bioluminescence detection.

β-galactosidase, encoded for an enzyme for which various chromogenic substrates are available.

β-glucuronidase (GUS) (Jefferson et al., EMBO J. 1987, 6, 3901-3907) or the uidA gene, which encodes an enzyme for various chromogenic substrates.

R-Locus gene product: protein that regulates the production of anthocyanin pigments (red coloring) in plant tissue and thus enables a direct analysis of the promoter activity without the addition of additional auxiliaries or chromogenic substrates (Dellaporta et al., In: Chromosome Structure and Function: Impact of New Concepts, 18th Stadler Genetics Symposium, 11: 263-282, 1988).

β-lactamase (Sutcliffe (1978) Proc Natl Acad Sei USA 75: 3737-3741), enzyme for various chromogenic substrates (eg PADAC, a chromogenic cephalosporin). xylE gene product (Zukowsky et al. (1983) Proc Natl Acad Sei USA 80: 1101-1105), catechol dioxygenase, which can convert chromogenic catechols.

Alpha amylase (Ikuta et al. (1990) Bio / technol. 8: 241-242).

Tyrosinase (Katz et al. (1983) J Gen Microbiol 129: 2703-2714), enzyme that oxidizes tyrosine to DOPA and dopaquinone, which consequently form the easily detectable melanin.

Aequorin (Prasher et al. (1985) Biochem Biophys Res Commun 126 (3): 1259-1268) can be used in calcium-sensitive bioluminescence detection.

The selection marker or the reporter gene is preferably encoded on the RE or RE / DSB construct and / or transformation construct, particularly preferably on the insertion sequence. However, it can also be encoded on an independent transformation construct which is introduced into the core or the plastids of a plant cell in a co-transformation with the transformation construct of interest.

The transformation vectors and insertion sequences according to the invention can contain further functional elements. The concept of further functional elements is to be understood broadly. Preference is given to all those elements which have an influence on the production, propagation, function, use or value of the insertion sequences, transformation constructs or vectors used in the process according to the invention. As examples, however, are not restrictive for the other functional elements:

i. Origins of replication (ORI; origin of DNA replication), which increase the expression cassettes or vectors according to the invention in, for example, E. coli or else in

Ensure plastids. Examples of E. coli ORI be mentioned are the pBR322 ori, the P15A ori (Sambrook et al... Molecular Cloning A Laboratory Manual, 2 ^nd ed Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, 1989) or the colEl- ORI, for example, from pBLUESCRIPT. Plastid ORIs are described in US 5,693,507, US 5,932,479 or WO 99/10513.

ii. Multiple cloning regions (MCS) allow and facilitate the insertion of one or more nucleic acid sequences. iii. Sequences that enable homologous recombination or insertion into the genome or piastome of a host organism.

iv. Elements for example "border sequences", which are an agrobacterium-mediated transfer in plant cells for the

Enables transmission and integration into the plant genome, such as the right or left border of the T-DNA or the vir region.

The introduction of an insertion sequence or an expression construct - for example for a recombinase, a DSBI enzyme or a selection marker - can advantageously be implemented using vectors into which these constructs or cassettes are inserted. Vectors can be, for example, plasmids, cosmids, phages, viruses, retroviruses or also agrobacteria.

In an advantageous embodiment, the expression cassette is introduced by means of plasmid vectors. Those vectors are preferred which ensure stable integration of the

Enable expression cassette into the host genome or plastom.

The production of a transformed organism or a transformed cell requires that the corresponding DNA is introduced into the corresponding host cell or into the plastids thereof. A large number of methods are available for this process, which is referred to as transformation (see also Keown et al. (1990) Methods in Enzymology 185: 527-537). For example, the DNA can be introduced directly by microinjection, electroporation or by bombardment with DNA-coated microparticles. The cell can also be chemically permeabilized, for example with polyethylene glycol, so that the DNA can get into the cell by diffusion. The transformation can also be carried out by fusion with other DNA-containing units such as minicells, cells, or lysosomes

Liposomes are done. Also to be mentioned are the transfection using calcium phosphate, DEAE-dextran or cationic lipids, transduction, infection, the incubation of dry embryos in DNA-containing solution, the sonication and the transformation of intact cells or tissue by micro or macro injection into tissue or embryos, tissue electroporation or the vacuum infiltration of seeds. Such processes are familiar to the person skilled in the art. In the case of injection or electroporation of DNA into plant cells, there are no special requirements for the plasmid used. Simple plasmids such as the pUC series can be used. If whole plants are to be regenerated from the transformed cells, it is useful that there is an additional selectable marker gene on the plasmid. The processes for regenerating plants from plant tissues or plant cells are also described. 5

There are various ways of introducing the DNA into the plastids. It is only crucial for the present invention that the DNA is introduced into the plastids. However, the present invention is not limited to one

10 specific procedure. Any method which allows the DNA to be transformed to be introduced into the plastids of a higher plant is suitable. The stable transformation of plastids is a process familiar to the person skilled in the art and has been described for higher plants (inter alia in Svab Z and Maliga P (1993) Proc

15 Natl Acad Sei USA 90 (3): 913-917). The methods are based, for example, on a transformation using a “particle gun” and an insertion into the plastid genome by homologous recombination under selection pressure. Further methods are described in US 5,877,402. In EP-A 0 251 654 the DNA is agro-

20 bacterium tumefaciens introduced (see De Block M et al. (1985) EMBO J 4: 1367-1372; Venkateswarlu K and Nazar RN (1991) Bio / Technology 9: 1103-1105). It has also been shown that DNA can be introduced into isolated chloroplasts by electroporation and thus transient expression can be achieved (To KY et al. (1996)

25 Plant J 10: 737-743). A transformation by means of a direct transfer of the DNA into plastids of protoplasts is preferred, for example using PEG (polyethylene glycol) (Koop HU et al. (1996) Planta 199: 193-201; Kofer W et al. (1998) In Vitro Cell Dev Biol Plant 34: 303-309; Dix PJ and Kavanagh TA (1995)

30 Euphytica. 85: 29-34; EP-A 0 223 247). Biolistic transformation methods are most preferred. The DNA to be transformed is e.g. Gold or tungsten particles applied. These particles are then accelerated to the explant to be transformed (Dix PJ and Kavanagh TA (1995) Euphytica.

35 85: 29-34; EP-A 0 223 247). Subsequently, transplasto plants are regenerated in a manner familiar to those skilled in the art under selection pressure on a suitable medium. Corresponding methods are described (e.g. US 5,451,513; US 5,877,402; Svab Z et al. (1990) Proc Natl Acad Sei USA 87: 8526-8530; Svab Z and Maliga P

40 (1993) Proc Natl Acad Sei USA 90: 913-917). In addition, the DNA can be introduced into the plastids by means of microinjection. A particular method of microinjection has recently been described (Knoblauch M et al. (1999) Nature Biotech 17: 906-909; van Bei AJE et al. (2001) Curr Opin Biotechnol

45 12: 144-149). This method is particularly preferred for the present invention. It is also possible to fuse the plastids from one species to another by protoplast fusion. bring to transform them there and then convert them back into the original way by protoplast fusion (WO 01/70939).

In addition to these "direct" transformation techniques, a transformation can also be carried out by bacterial infection using Agrobacterium tumefaciens or Agrobacterium rhizogenes (Horsch RB (1986) Proc Natl Acad Sei USA 83 (8): 2571-2575; Fraley et al. (1983) Proc Natl Acad Sei USA 80: 4803-4807; Bevans et al. (1983) Nature 304: 184-187). The expression cassette, for example for the DSBI enzyme, is preferably integrated into special plasmids, either into a shuttle or intermediate vector or a binary vector. Binary vectors are preferably used. Binary vectors can replicate in both E. coli and Agrobacterium and can be transformed directly into Agrobacterium (Holsters et al. (1978) Mol Gen Genet 163: 181-187). Various binary vectors are known and some are commercially available, for example pBIN19 (Clontech Laboratories, Inc. USA; Bevan et al. (1984) Nucl Acids Res 12: 8711). The selection marker gene allows selection of transformed agrobacteria and is, for example, the nptll gene which confers resistance to kanamycin. The binary plasmid can be transferred into the agrobacterial strain, for example, by electroporation or other transformation methods (Mozo & Hooykaas 1991, Plant Mol. Biol. 16, 917-918). The coculture of the plant explants with the agrobacterial strain usually takes place for two to three days. The agrobacterium, which acts as the host organism in this case, should already contain a plasmid with the vir region. Many strains of Agrobacterium tumefaciens are able to transfer genetic material, e.g. the strains EHAl01 [pEHAl01] (Hood EE et al. (1996) J Bacteriol 168 (3): 1291-1301), EHA105 [pEHA105] (Hood et al. (1993) Transgenic Research 2: 208-218), LBA4404 [ pAL4404] (Hoekema et al. (1983) Nature 303: 179-181), C58Cl [pMP90] (Koncz and Schell (1986) Mol Gen Genet 204: 383-396) and C58Cl [pGV2260] (Deblaere et al. (1985 ) Nucl Acids Res 13: 4777-4788).

For the transfer of the DNA to the plant cell, plant explants with Agrobacterium tumefaciens or Agrobacterium rhizogenes are co-cultivated. Starting from infected plant material (e.g. leaf, root or stem parts, but also protoplasts or suspensions of plant cells), whole plants can be regenerated using a suitable medium, which can contain, for example, antibiotics or biocides for the selection of transformed cells. A co-transformed selection marker allows the selection of transformed cells from untransformed (McCormick et al. (1986) Plant Cell Reports 5: 81-84). The plants obtained can be grown, grown and crossed in a conventional manner. Two or more generations should be cultivated to ensure that genomic integration is stable and inheritable. The methods mentioned are described for example in Jenes B et al. (1993) Techniques for Gene Transfer, in: Transgenic Plants, Vol. 1, Engineering and Utilization, edited by Kung SD and Wu R, Academic Press, pp. 128-143 and in Potrykus (1991) Ann Rev Plant Physiol Plant Molec Biol 42: 205-225).

The Agrobacterium-mediated transformation is best suited for dicotyledonous plant cells, whereas the direct transformation techniques are suitable for every cell type.

The Agrobacterium-mediated transformation is particularly preferably used for the core transformation, the direct transformation techniques are particularly preferably used for the plastid transformation.

Once a predominantly homotransplastoma plant cell has been produced by the method according to the invention, a complete plant can be obtained using methods known to the person skilled in the art. This is based, for example, on callus cultures. The formation of shoots and roots can be induced in a known manner from these still undifferentiated cell masses. The sprouts obtained can be planted out and grown.

deletion method

In the described methods according to the invention, it is advantageous at various stages to again remove certain previously introduced sequences (for example for selection markers, recombinases and / or DSBI enzymes) from the piastome or genome

Remove plant or cell. It is advantageous but not absolutely necessary to remove the selection marker, which was introduced, for example, when inserting a non-natural RE sequence and / or DBS recognition sequence, from the master plant, because then in a subsequent transformation (eg with the insertion sequence) the same selection marker can be used again. A deletion is particularly advantageous since the selection marker is no longer absolutely necessary after the selection phase and is therefore superfluous. The deletion also increases consumer acceptance and is desirable considering regulatory considerations. In addition, the plastid's protein synthesis apparatus does not become unnecessary burdened by the synthesis of the marker protein, which has a potentially advantageous effect on the properties of the corresponding plant.

Various methods are known to the person skilled in the art for the deletion of sequences. However, excision by means of recombinases should not be mentioned as a limitation. Various sequence-specific recombination systems are described, such as the Cre / lox system of Bacteriophagen Pl (Dale EC and Ow DW (1991) Proc Natl Acad Sei USA 88: 10558-10562; Russell SH et al. (1992) Mol Gen Genet 234: 49- 59; Osborne BI et al. (1995) Plant J. 7, 687-701), the yeast FLP / FRT system (Kilby NJ et al. (1995) Plant J 8: 637-652; Lyznik LA et al. ( 1996) Nucleic Acids Res 24: 3784-3789), the Mu Phage gin recombinase, the E. coli pin recombinase or the R / RS system of the pSRI plasmid (Onouchi H et al. (1995) Mol Gen Genet 247: 653 -660; Sugita K et al. (2000) Plant J 22: 461-469). These methods can be used not only for the deletion of DNA sequences from the nuclear genome, but also from the piastome (Corneille et al. (2001) Plant J 27: 171-178; Hajdukieicz et al. (2001) Plant J

27: 161-170). Other recombinases that can be used are, for example: PhiC31 (Kuhstoss & Rao (1991) J Mol Biol 222: 897-908), TP901-1 (Christiansen et al. (1996) J Bacteriol 178: 5164-5173), xisF from Anabaena ( Ramaswamy et al. (1997) Mol Microbiol 23: 1241-1249), PhiLC3 integrase (Lillehaug et al. (1997) Gene 188: 129-136) or the recombinase encoded by the R4 phage gene (Matsuura et al (1996) J Bacteriol 178: 3374-3376).

In a preferred embodiment, however, the deletion is realized by intramolecular (eg intrachromosomal) recombination on the basis of sequence duplications introduced accordingly. The efficiency of the latter can be increased by the targeted introduction of double-strand breaks near the sequence duplications (cf. FIG. 8). For this purpose, the sequence to be deleted is flanked on both sides by homology sequences H1 and H2, which are of sufficient length and homology to recombine with one another. The recombination is induced by inducing at least one sequence-specific double-strand break near one of the two homology sequences located in the DSB recognition sequence. This DSB recognition sequence is preferably located between the two homology sequences. A DSBI enzyme is preferably expressed or introduced to induce the double-strand break. This method is particularly preferably used for the deletion of selection markers from the piastome. Another object of the invention relates to the transplastomic, predominantly homoplastic, plants produced by the process according to the invention, and parts thereof, such as leaves, roots, seeds, fruits, tubers, pollen or cell cultures, callus, etc. - derived from such.

Another object of the invention relates to the plants used in the method according to the invention which have an expression cassette according to the invention for a fusion protein from a PLS and a recombinase of the family of

Resolvases / invertases under the control of a promoter functional in the plant cell nucleus. The expression cassette is particularly preferably stably integrated into the nuclear DNA. Also included are parts of the same, such as leaves, roots, seeds, tubers, fruits, pollen or cell cultures, callus, etc. - derived from such plants.

The invention further relates to expression cassettes which contain nucleic acid sequences coding for recombinases of the resolvase / invertase family under the control of a promoter which is functional in plant plastids. The expression cassette is particularly preferably stably integrated into the piastom. Also included are parts of the same, such as leaves, roots, seeds, tubers, fruits, pollen or cell cultures, callus, etc. - derived from such plants.

Genetically modified plants according to the invention which can be consumed by humans and animals can also be used, for example, directly or after preparation known per se as food or feed.

The invention further relates to the use of the transplastomic, predominantly homoplastic, plants according to the invention described above and the cells, cell cultures, parts derived from them, such as roots, leaves, etc. in transgenic plant organisms, and transgenic propagation material such as seeds or fruits, for the production of food or feed, pharmaceuticals or fine chemicals.

Fine chemicals means enzymes such as the industrial enzymes mentioned below, vitamins such as tocopherols and tocotrienols (eg vitamin E) and vitamin B2, amino acids such as methionine, lysine or glutamate, carbohydrates such as starch, amylose, amylopectin or sucrose, fatty acids such as for example saturated, unsaturated and polyunsaturated fatty acids, natural and synthetic flavors, aromas such as linalo- 01. menthol, borneon (camphor), pinene, limonene or geraniol and dyes such as retinoids (e.g. vitamin A), flavonoids (e.g. quercetin, rutin, tangeretin, nobiletin) or carotenoids (e.g. ß-carotene, lycopene, astaxanthin). The production of tocopherols and tocotrienols and carotenoids is particularly preferred. The transformed host organisms and the isolation from the host organisms or from the growth medium are cultivated using methods known to those skilled in the art. The production of pharmaceuticals such as antibodies or vaccines has been described (Hood EE, Jilka JM. (1999) Curr Opin Biotechnol. 10 (4): 382-386; Ma JK and Vine ND (1999) Curr Top Microbiol Immunol .236 : 275-92).

The method according to the invention is particularly suitable for the production of industrial enzymes in the context of so-called "phytofarming". Examples, however, are not restrictive for the industrial enzymes: lipases, esterases, proteases, nitrilases, acylases, epoxyhydrolases, amidases, phosphatases, xylanases, alcohol dehydrogenases, amylases, glucosidases, galactosidases, pullulanases, endocellulases, glucancleas, nuclitaminases, cellenase, cellulase, cellulase, cellulase, cellulase, cellulase, monolitase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, cellulase, and cellulase type , Lysozymes and laccases.

Embodiments which are particularly preferred in the context of the invention are described in more detail below in the context of the explanations for the figures.

sequences

1. SEQ ID NO: 1 pCB42-94 base vector for plastid transformation.

2. SEQ ID NO: 2

Nucleic acid sequence inserted into the multiple cloning site of pCB42-94 (SEQ ID NO: 1). Resulting vector: PCB199-3.

3. SEQ ID NO: 3

Nucleic acid sequence inserted into the multiple cloning site of pCB42-94 (SEQ ID NO: 1). Resulting vector: PCB401-20

4. SEQ ID NO: 4

Synthetic sequence of the homing endonuclease I-Ppol (ORF: 16 to 507) 5. SEQ ID NO: 5

Protein sequence of the homing endonuclease I-Ppol

6. SEQ ID NO: 6 oligonucleotide primers pl9 5 5 '-TAAGGCCCTCGGTAGCAACGG-3'

7. SEQ ID NO: 7 oligonucleotide primer p20 5 '-GGGGTACCAAATCCAACTAG-3'

10. SEQ ID NO: 8 oligonucleotide primers p21: 5 '-GGAGCTCGCTCCCCCGCCGTCGTTC-3'

9. SEQ ID NO: 9 oligonucleotide primer p22 5 '-GATGCATGATGACTTGACGGCATCCTC-3'

15

10. SEQ ID NO: 10

Nucleic acid sequence coding for the transit peptide of the small subunit (SSU) of ribulose bisphosphate carboxylase (Rubisco ssu) from pea 20

11. SEQ ID NO: 11

Transit peptide of the small subunit (SSU) of pea ribulose bisphosphate carboxylase (Rubisco ssu)

25 12. SEQ ID NO: 12

Transition peptide of tobacco plastid transketolase.

13. SEQ ID NO: 13

Nucleic acid sequence coding for the transit peptide of 30 plastid transketolase from tobacco (reading frame 1; pTP09)

14. SEQ ID NO: 14

Nucleic acid sequence coding for the transit peptide of tobacco plastid transketolase (reading frame 2; pTPlO) 35

15. SEQ ID NO: 15

Nucleic acid sequence coding for the transit peptide of tobacco plastid transketolase (reading frame 3; pTPll)

40 16. SEQ ID NO: 16

Arabidopsis thaliana plastid isopentenyl pyrophosphate isomerase-2 (IPP-2) transit peptide.

45 17th SEQ ID NO: 17

Nucleic acid sequence coding for the transit peptide of the plastid isopentenyl pyrophosphate isomerase-2 (IPP-2) from Arabidopsis thaliana (reading frame 1; IPP-9) 5

18. SEQ ID NO: 18

Nucleic acid sequence coding for the transit peptide of the plastid isopentenyl pyrophosphate isomerase-2 (IPP-2) from Arabidopsis thaliana (reading frame 2; IPP-10) 10

19. SEQ ID NO: 19

Nucleic acid sequence coding for the transit peptide of the plastid isopentenyl pyrophosphate isomerase-2 (IPP-2) from Arabidopsis thaliana (reading frame 3; IPP-11) 15

20. SEQ ID NO: 20

Nucleic acid sequence coding for the PrbcL promoter from tobacco.

20 21. SEQ ID NO: 21

Nucleic acid sequence coding for the tobacco Prpsl6-107 promoter.

22. SEQ ID NO: 22

25 nucleic acid sequence coding for the Prrnlδ promoter from tobacco.

23. SEQ ID NO: 23

Nucleic acid sequence coding for promoter PaccD-129 from 30 tobacco.

24. SEQ ID NO: 24

Nucleic acid sequence coding for the promoter PclpP-53 from tobacco. 35

25. SEQ ID NO: 25

Nucleic acid sequence coding for the promoter Prrn-62 from tobacco.

40 26. SEQ ID NO: 26

Nucleic acid sequence coding for the promoter Prpsl6 from tobacco.

27. SEQ ID NO: 27 45 nucleic acid sequence coding for the promoter PatpB / E-290 from tobacco. 28 SEQ ID NO: 28

Nucleic acid sequence coding for the promoter PrpoB-345 from tobacco.

5 29 SEQ ID NO: 29

Nucleic acid sequence coding for a sequence derived from the consensus sequence of the σ70 promoters from E. coli.

30. SEQ ID NO: 30

10 nucleic acid sequence coding for the 5 'untranslated region of the psbA gene from tobacco (5'psbA)

31. SEQ ID NO: 31

Nucleic acid sequence coding for the 5 'untranslated 15 region including 5' parts from the coding region of the rbcL gene from tobacco (5'rbcL).

32. SEQ ID NO: 32

Nucleic acid sequence coding for the 5 'untranslated 20 region of the rbcLs gene from tobacco.

33. SEQ ID NO: 33

Nucleic acid sequence coding for the 3 'untranslated region of the psbA-1 gene from Synechocystis (3'psbA-1) 25

34. SEQ ID NO: 34

Nucleic acid sequence coding for the 3 'untranslated region of the psbA gene from tobacco (3'psbA)

30 35. SEQ ID NO: 35

Nucleic acid sequence coding for the 3 'untranslated region of the rbcL gene from tobacco (3'rbcL)

36. SEQ ID NO: 36

35 nucleic acid sequence coding for synthetic ribosome binding sites (RBS)

37. SEQ ID NO: 37

Nucleic acid sequence coding for synthetic ribosome binding site (RBS)

38. SEQ ID NO: 38 oligonucleotide primer p65: 5 '-ATGGATCCATATGGCCATGGCACAAGGGGTTGTG-3'

45 39. SEQ ID NO: 39 oligonucleotide primer p66: 5 '-CCCGGGCTCGAGCTGCAGCTACGCCGCTACGTC -3' 40. SEQ ID NO: 40

Insert of vector pCBl93-17 comprising expression cassette coding for fusion protein from PLS and ΦC31 recombinase

5 41. SEQ ID NO: 41

Amino acid sequence of the fusion protein from PLS and ΦC31 recombinase

42. SEQ ID NO: 42

10 amino acid sequence of the ΦC31 recombinase

43. SEQ ID NO: 43 oligonucleotide primer p71: 5 '-TGGATCCATATGGCCATGGCTAAGAAAGTAG-3'

15 44. SEQ ID NO: 44 oligonucleotide primer p72: 5 '-CCTCGAGCTGCAGTTAAGCGAGTTGG-3'

45. SEQ ID NO: 45

Binary vector pCB245-37 comprising expression cassette coding 20 for fusion protein from PLS and TP901-1 recombinase

46. SEQ ID NO: 46

Amino acid sequence of the fusion protein from PLS and TP901-1 recombinase 25

47. SEQ ID NO: 47

Amino acid sequence of the TP901-1 recombinase

48. SEQ ID NO: 48

30 vector pCB249-24 comprising attP recognition region for the recombinase from TP901-1 and attP recognition region for the recombinase from ΦC31

49. SEQ ID NO: 49

35 vector pCB384-40 comprising attP recognition region for the recombinase from TP901-1 and attP recognition region for the recombinase from ΦC31 and an expression cassette for the homing endonuclease I-Ppol

40 50th SEQ ID NO: 50 nucleic acid sequence coding for the insert from vector pCB456-2

51. SEQ ID NO: 51 oligonucleotide primer p354

5 '-CGGATCCGGAGGCATCATGACTAAGAAAGTAGCAATCTATACACG 45 52. SEQ ID NO: 52 oligonucleotide primer p355

5 '-ATAAGCTTTATTCTCGATATCG-3'

53. SEQ ID NO: 53 nucleic acid sequence coding for insert 5 from vector pCB557-l

54. SEQ ID NO: 54 nucleic acid sequence coding for the insert from vector pCB554-3

10 55. SEQ ID NO: 55 oligonucleotide primer p268

5 '-TCGACTTGACATTCACTCTTCAATTATCTATAATGATACATG-3'

56. SEQ ID NO: 56 oligonucleotide primer p72

5 '-CCTCGAGCTGCAGTTAAGCGAGTTGG-3' 15

57. SEQ ID NO: 57 nucleic acid sequence coding for PCR product

TP-Rek

58. SEQ ID NO: 58 nucleic acid sequence coding for Southern 0 blot probe

59. SEQ ID NO: 59 nucleic acid sequence coding for one

Promoter sequence derived from the consensus sequence of the σ70 promoters 5 from E. coli.

60. SEQ ID NO: 60 nucleic acid sequence coding for the insert from vector pCB487-20 comprising sequence coding for TP901 recombinase 0

61. SEQ ID NO: 61 amino acid sequence coding for TP901

recombinase

Figures 5

In the context of the method according to the invention, the embodiments illustrated in the following figures are particularly preferred. The following abbreviations generally apply to the illustrations: 0

X / Y: Recombinase recognition parts

U / W: Recombinase recognition parts, preferably cannot recombine with X / Y 5 X '/ Y': Recombinase recognition site, can preferably only recombine with X / Y

U '/ W': Recombinase recognition site, can preferably only recombine with U / W

X / Y 'or X' / Y: emerged after recombination between X / Y and X '/ Y', between these sites the recombinase alone can preferably not catalyze any further recombinations

U / W 'or U' / W: emerged after recombination between U / W and U '/ W', between these sites the recombinase alone can preferably not catalyze any further recombinations

R, Rl, R2: recombinase

A, A 'pair of homologous sequences A and A'

A / A 'Result of a homologous recombination between A and A' and / or an exchange between A and A 'due to repair synthesis.

B, B 'pair of homologous sequences B and B'

B / B 'Result of a homologous recombination between B and B' and / or an exchange between B and B 'due to repair synthesis.

Hl, H2: pair of homologous sequences Hl and H2

Hl / 2: sequence as a result of the homologous recombination of Hl and H2

DS functional DSB recognition sequence

nDS non-functional "half-page" of a DSB recognition sequence

E: DSBI enzyme

P, P ': promoter

I: further nucleic acid sequence S, S 'positive selection markers

NS negative selection marker

1. Fig.LA/B: Introduction of a RE sequence in the piastom by means of double "cross-over"

In a particularly preferred embodiment 1, an RE construct is first introduced into plastids of a higher plant. In this embodiment, the RE construct is preferably equipped with homologous target regions and with an expressible selection marker (promoter - 5'UTR - selection marker - 3'UTR) and in this embodiment preferably contains a RE sequence. The RE construct can - optionally - already encode further genes of interest. Mostly homoplastic master plants are produced.

As already described above, A / A 'or B / B' represents the result of a homologous recombination. Even if these sequences are still present in the piastom in subsequent steps, they are no longer shown in the corresponding figures, since they are functionally irrelevant are.

2. Fig. 2: Introduction of an insertion sequence with an expression cassette for a recombinase and possibly

Selection markers and other genes of interest

Explants of the master plants produced by embodiment 1 are used for a further transformation with a circular transformation construct according to the invention. The recombinase is expressed by the transformation construct itself under the control of a promoter that is functional in plastids.

3. Fig.3: Introduction of an insertion sequence with a selection marker, if necessary, and other genes of interest and expression / introduction of the recombinase into trans

In a further preferred embodiment, the recombinase is not encoded by the transformation construct, but rather is either expressed in trans (in plastids or as a PLS fusion protein in the core) or in the form of RNA or as protein in the plastids. This embodiment is particularly preferred when the transformation construct does not comprise any promoter elements and an expression of the encoded genes only after insertion. tion into the piastome using plastidic, endogenous promoters.

4. Fig.4: Introduction of a RE construct with two RE sequences

In a further embodiment 2, an RE construct with two RE sequences is first introduced into plastids of a higher plant. In this embodiment, the RE construct is preferably equipped with homologous target regions and with an expressible selection marker (promoter - 5'UTR - selection marker - 3'UTR). The two introduced RE sequences may be the same or different, but preferably cannot recombine with one another under the influence of a recombinase.

The RE construct can - optionally - already encode further genes of interest. Mostly homoplastic master plants are produced.

5. Fig. 5: Introduction of an insertion sequence with two RE sequences and possibly with selection markers and other genes of interest

Explants of the master plants produced under embodiment 2 (cf. FIG. 4) are used for a further transformation with a transformation construct according to the invention, which in turn has two RE sequences. The construct can be linear or circular (Fig. 5). The two RE sequences may be the same or different, but preferably cannot recombine with one another under the influence of a recombinase.

The recombinase or the recombinases can be expressed by the transformation construct itself under the control of a promoter which is functional in plastids or else can be expressed in trans (in plastids or as a PLS fusion protein in the core) or in the form of RNA or as a protein in the plastids are transfected.

6. Fig. 6: Introduction of a RE / DSB construct with a RE sequence and a DSB recognition sequence

In a further embodiment 3, a RE / DSB construct with at least one RE sequence and one DSB recognition sequence is first introduced into plastids of a higher plant. In this embodiment, the RE construct is preferably homologous Target regions and equipped with an expressible selection marker (promoter - 5'UTR - selection marker - 3'UTR).

The RE construct can - optionally - already encode other genes of interest - such as a negative selection marker. Mostly homoplastic master plants are produced.

7. Fig. 7A-C: Introduction of an insertion sequence and expression / introduction of a recombinase and a DSBI enzyme

Explants of the master plants produced according to embodiment 3 (cf. FIG. 6) are used for a further transformation with a transformation construct according to the invention, which in turn has a RE sequence and additionally contains sequence sections which are homologous to those in the RE / DSB construct (e.g. Sequence of a selection marker or a part thereof). A corresponding recombination product results from the action of a recombinase (FIG. 7A).

The recombinase can be expressed by the transformation construct itself under the control of a promoter which is functional in plastids or else can be expressed in trans (in plastids or as a PLS fusion protein in the nucleus) or in the form of RNA or as protein in the plastids.

The recombination product is then exposed to the action of a DSBI enzyme that recognizes and cuts the DSB recognition sequence introduced by the RE / DSB construct. The DSBI enzyme can be expressed by the transformation construct itself under the control of a promoter which is functional in plastids or else can be expressed in trans (in plastids or as a PLS fusion protein in the nucleus) or in the form of RNA or as a protein in the plastids be transfected. The cut creates an intramolecular homologous recombination between homologous sequences. In the case shown, these homologous sequences are represented, for example, by duplicating a selection marker or parts thereof. The homologous recombination causes the deletion of the DSB recognition sequence. In a preferred embodiment, the homologous recombination additionally deletes a negative selection marker (FIG. 7B).

The piastomart produced in this way can no longer be cut by the DSBI enzyme which is still present, while the copies of the master plant produced with the RE / DSB construct, which have not yet inserted the insertion sequence by recombination, can still be cut. This builds in Selection pressure, which - supported by a repair synthesis - leads to a rapid accumulation of the desired plastid DNA molecules (Fig. 7C).

Those skilled in the art are aware that the order of the expressed genes in an operon is interchangeable and can vary in the embodiments described above. In principle, the recombinase and / or the DSBI enzyme can be expressed on the transformation construct and / or separately (in the core or plastids) or introduced in another way - for example by transfection with RNA or protein in plastids.

8. Fig. 8: Deletion of sequences by means of intramolecular homologous recombination induced by sequence-specific double-strand breaks

In all of the above-described embodiments, sequences — for example coding for selection markers or DSBI enzymes — are preferably flanked by homology sequences Hl and H2 which are of sufficient length and homology to recombine with one another. The recombination is induced by induction of at least one double-strand break between the two homology sequences located in the DSB recognition sequence. A DSBI enzyme is preferably transiently expressed or introduced to induce the double-strand break (FIG. 8).

Those skilled in the art are aware that the order of the expressed genes in an operon is interchangeable and can vary in the embodiments described above. If only one homology sequence is used to insert the insertion sequence, it can also be located on the 5 '(as shown in the figures) or 3' side of the double-strand break. In principle, the DSBI enzyme can be expressed on the transformation construct and / or separately (in the core or plastids) or introduced in another way - for example by transfection with RNA or protein in plastids.

9. Fig. 9: Southern analysis of predominantly homotransplastomas

plants

Wild-type and predominantly homotransplastome master plants were analyzed for the modification (introduction of a RE sequence) (cf. Example 4). The modification detected a band of 1750 bp (lanes 2, 3, and 4 corresponding to lines CB199NTH-4, -6, and -8), while a band of 3100 bp was detected in the unmodified wild type plant (lane 1) , Fig. 10: Southern analysis of predominantly homotransplastomic plants

A: Wild-type and predominantly homotransplastome master plants were analyzed for the modification (introduction of a RE and DSB recognition sequence) in a Southern blot (cf. Example 4). As a result of the modification, a band of approximately 3.8 kb was detected in the DNA treated here with HindIII (lanes 2, 3 and 4 corresponding to lines CB199NTH-4, -6, and -8), while a band was found in the unmodified wild-type plant of about 7.7 kb was detected (lane 5). In lines CB199NTH-4, -6, and -8, in addition to the expected approximately 3.8 kb band and the approximately 7.7 kb band, a band at approximately 5.2 kb can also be seen. This band results (as illustrated in B) from the hybridization of the 16SrRNA promoter which is contained in the probe with the 16SrRNA promoter which can be found downstream of the selection marker aadA. This approximately 5.2 kb band is significantly weaker than the 3.8 kb band because only a small proportion of the probe actually hybridizes with this fragment (wt, unmodified wild type plant; m - molecular size standard)

B: Schematic representation of the HindIII fragments which were to be expected in A by hybridization with the probe (probe) in an unmodified wild type plant (top) and a modified plant CB199NTH (bottom). (trnV - gene coding for a tRNA-Val; rrnl6 - gene coding for the 16SrRNA; aadA - gene coding for a selection marker; 3 'psbA - non-coding region downstream of the psbA gene, here incorporated into the expression cassette for the selection marker aadA)

Fig. 11: Southern analysis of predominantly homotransplastomic plants

A: Wild-type and predominantly homotransplastome master plants were analyzed for the modification (introduction of a RE and DSB recognition sequence) in a Southern (cf. Example 14). As a result of the modification, a band of approximately 1.7 kb was detected in the DNA treated here with EcoRI (lanes 1 and 4 corresponding to lines CB456NTH-1 and -15), while a band of approximately 3.1 kb was found in the unmodified wild-type plant was detected (lane 6). (wt - unmodified wild type plant; wild type - shows the expected fragment size in unmodified wild type plants; trans- genic - shows the expected fragment size in plants CB456NTH)

B: Schematic representation of the EcoRI fragment which was to be expected in A by hybridization with the probe (probe) in a modified plant CB456NTH. (trnV - gene coding for a tRNA-Val; rrnl6 - gene coding for the 16SrRNA; aadA - gene coding for a selection marker; 3 psbA (Synec) - non-coding region downstream of the psbA-1 gene from Synechocystis, incorporated here in the

Expression cassette for the selection marker aadA; Psynth. - synthetic promoter derived from the consensus sequence for E. coli σ70 promoters).

Examples

General methods:

The chemical synthesis of oligonucleotides can be carried out, for example, in a known manner using the phosphoamidite method (Voet, Voet, 2nd edition, Wiley Press New York, pages 896-897). The cloning steps carried out in the context of the present invention, such as e.g. Restriction cleavages, agorose gel electrophoresis, purification of DNA fragments, transfer of nucleic acids to nitrocellulose and nylon membranes, linking of DNA fragments, transformation of E. coli cells, cultivation of bacteria, multiplication of phages and sequence analysis of recombinant DNA are like Sambrook et al. (1989) Cold Spring Harbor Laboratory Press; ISBN 0-87969-309-6. The sequencing of recombinant DNA molecules is carried out with a laser fluorescence DNA sequencer ALF-Express (Pharmacia, Upsala, Sweden) according to the method of Sanger (Sanger et al. (1977) Proc Natl Acad Sei USA 74: 5463-5467).

Example 1: Creating a basic vector for plastid transformation

First, the selected target regions were cloned from the piastome of the tobacco variety SRI using PCR. The left target region was amplified with the primers pl9 and p20.

pl9: 5 '-TAAGGCCCTCGGTAGCAACGG-3' (SEQ ID NO: 6)

p20: 5 '-GGGGTACCAAATCCAACTAG-3' (SEQ ID NO: 7) The primers p21 and p22 were used to amplify the right target region, with the latter primer also introducing spectinomycin resistance into the amplified part of the 16SrDNA in addition to the SRI resistance (binding type marker).

p21: 5 '-GGAGCTCGCTCCCCCGCCGTCGTTC-3' (SEQ ID NO: 8)

p22: 5 '-GATGCATGATGACTTGACGGCATCCTC-3' (SEQ ID NO: 9)

The two amplified regions were cloned into pBluescript and pZeroBlunt and sequenced. The left and right target regions were then cloned into the backbone of the pUC19 vector. The interfaces EcoO109l and PvuII of the vector were used for this. A multiple cloning site from the pBluescript (from Kpnl to Sacl) was cloned between the left and right target regions. This is located in the base vector for the plastid transformation between the two plastidically encoded genes trnV and rrnl6. This base vector for plastid transformation was given the designation pCB42-94 (SEQ ID NO: 1). The vector contains the following sequence elements:

a) Position complementary bp 55-1405: right target region with the partial gene of 16SrRNA (complementary bp 56 to 1322). In the latter there are mutations for streptomycin resistance (SRI, position bp 346) and spectinomycin resistance (SPC1, position bp 68).

b) Position complementary bp 2374 to 1510: left target region including ORF131 (bp 1729 to 2124) and trnV gene

(complementary bp 1613 to 1542).

c) Position bp 1404 to 1511: multiple cloning site

d) Position bp 2629 to 3417: ampicilin resistance in the vector backbone

Example 2: Creation of a vector (pCB199-3) for the introduction of a non-naturally occurring DSB recognition sequence for the homing endonucleases

I-Ppol and RE sequences (attB) for the recombinases ΦC31 and TP901-1 in the piastome of tobacco

Various elements were successively cloned into the multiple cloning site of the base vector pCB42-94 (SEQ ID NO: 1) for the plastid transformation: a) for recognition region (mutated, does not contain an Xbal interface; complementary 1307-1354)

b) Expression cassette for the expression of the marker gene aadA consisting of:

i) Promoter of the gene for 16SrRNA (complementary 1191-1281)

ii) 5 'untranslated region of the tobacco rbcL gene ^(com-' tary 1167-1184) including mutated 5 'portions of the rbcL gene (6 AS duplication of the rbcL gene, partially mutated, as a result of the cloning strategy, so that a fusion encoding a total of 12 amino acids (complementary 1131-1166) with the following element, the aadA gene, was created)

iii) aadA gene (complementary 336-1130)

iv) the 3 'region of the psbA gene (complementary 232-323)

c) attB RE sequences for the recombinases ΦC31 (position base pair 131 to 170) and TP901-1 (position base pair 21 to 121)

d) Core recognition region for the homing endonuclease I-Ppol (position base pair complementary 176 to 190).

Instead of the multiple cloning site in the base vector for plastid transformation, this vector with the designation pCB199-3 contains the elements listed within the framework of the nucleic acid sequence with SEQ ID NO: 2. The sequence which replaces the complete MCS from Kpnl to Sacl is given. However, due to the cloning strategy, there is no longer a Kpnl interface in the sequence given.

Example 3: Creation of a further vector (pCB401-20) for the introduction of a non-naturally occurring recognition region for the homing endonucleases I-Ppol and RE sequences (attB) for the recombinases ΦC31 and TP901-1 in the piastome of tobacco

In contrast to the vector pCB199-3 described in Example 2, the vector described here contains no promoter and 3'UTR, which was linked directly to the selection marker aadA. Rather, the expression of the aadA gene is controlled starting from the promoter of the trnV gene, which is located in the plastid genome or in the left target region upstream of the aadA gene. The aim of creating this vector was to duplicate sequences to avoid cations by exploiting regulatory areas from the tobacco plastid genome. For this purpose, various elements were successively cloned into the multiple cloning site of the base vector pCB42-94 for the plastid transformation:

a) Ribosome binding site (complementary bp 1033 to 1050)

b) aadA gene (complementary bp 238 to 1032)

c) attB RE sequences for the recombinases ΦC31 (position base pair 131 to 170) and TP901-1 (position base pair 21 to 121)

d) core recognition region for the homing endonucleases I-Ppol (complementary bp 176 to 190)

The vector obtained in this way also conferred spectinomycin resistance in E. coli. Instead of the multiple cloning site in the base vector pCB42-94 (SEQ ID NO: 1) for the plastid transformation, this vector with the designation pCB401-20 contains the elements listed within the framework of the nucleic acid sequence with the SEQ ID NO: 3. Here, too, the whole is the MCS replacing sequence (from Sacl to Kpnl) given.

Example 4: Creating predominantly homoplastomic tobacco master plants that have a non-natural DSB

Recognition sequence and RE sequences (attB) included

The plasmid pCB199-3 was introduced into the plastids of tobacco (Nicotiana tabacum cv. Petit Havana) as described below. The regenerated plants were named CB199NTH. Independent lines have been given different end numbers (e.g. CB199NTH-4).

Analogously, the vector pCB401-20 is introduced into the plastids of tobacco. The resulting plants are designated CB401NTH accordingly.

First, leaf disks with a diameter of 2.0 to 2.5 cm were cut out from plants grown in vitro using a sterile cork borer and the top of each leaf was placed on a petri dish with bombardment medium (MS salts (Sig a-Aldrich): 4.3 g / L; sucrose: 30.0 g / L, phytoagar (Duchefa, P1003): 0.6% (w / v), pH 5.8 After autoclaving, 1.0 mg / L thiamine (Duchefa, T0614 ) and 0.1 g / L myo-innositol (Duchefa, 10609) added). The underside of the leaf facing away from the agar was then bombarded with the particle gun. For this purpose, the plasmid DNA to be transformed (purified from E. coli by means of Nucleobond AX100 Macherey & Nagel) applied according to the following protocol to 0.6 μ gold particles ("coating"). First, 30 mg of gold powder (BioRad) was taken up in ethanol. 60 μl of the gold suspension are transferred to a new Eppendorf tube and the gold particles are sedimented by centrifugation (for 10 seconds). The gold particles were washed twice in 200 μl sterile water and, after a further centrifugation step, taken up in 55 μl water. With constant mixing (vortexing), the following was quickly added: 10

5 μl plasmid DNA (1 μg / μl)

50 ul 2.5 M CaCl ₂

20 ul 0.1 M spermidine

The suspension was then vortexed for a further 3 minutes and then briefly centrifuged. The sedimented gold-DNA complexes were washed 1 to 2 times in 200 μl ethanol and after a further centrifugation step they were finally taken up in 63 μl ethanol. 3.5 μl (shot

20 speaking 100 μg gold) of this suspension applied to a macro carrier.

According to the manufacturer's instructions, the particle cannon (BioRad, PDSlOOOHe) was prepared and the leaf explants at a distance

Bombarded 25 by 10 cm with the gold-DNA complexes. The following parameters were used: Vacuum: 27 inch Hg, pressure 1100 psi. After the bombardment, the explants are incubated for 2 days in climatic chambers (24 ° C, 16 h light, 8 h dark) and then divided into segments of approximately 0.5 cm2 using a scalpel. This

30 segments were then placed on regeneration medium [bombardment medium plus 1 mg / L 6-benzylaminopurine (BAP, Duchefa, B0904) and 0.1 mg / L naphthylacetic acid (NAA, Duchefa, N0903)] plus 500 mg / L spectinomycin (Duchefa, S0188 ) transferred and incubated for 10 to 14 days in the above-mentioned conditions in a climatic chamber.

35 At the end of this time, the leaf segments were transferred to fresh regeneration medium plus 500 mg / L spectinomycin. This process was repeated until green shoots formed on the explants. The rungs were separated with a scalpel and on growth medium (like bombardment medium, however

40 10 g / L sucrose instead of 30 g / L sucrose) plus 500 mg / L spectinomycin.

Optionally, in order to obtain predominantly homoplastomic plants, explants 45 can again be cut off from the regenerated plants and placed on regeneration medium with 1000 mg / L spectinomycin. Regenerating shoots are placed in jars with growth medium plus 500 to 1000 mg / L spectinomycin. After rooting, the plants are transferred to the greenhouse and grown there until they reach seed maturity.

In the case of transformation with the plasmid pCB199-3, 8 plants were obtained which showed resistance to spectinomycin. Using PCR and Southern analyzes, it was demonstrated that three of these lines (lines CB199NTH-4, -6 and -8) actually have incorporated the aadA gene into the plastid genome.

To analyze the transplastomic plants according to Southern, total DNA from leaves of transformed and non-transformed plants was isolated using the GenElute Plant Genomic DNA Kit (Sigma). The DNA was taken up in 200 ul eluate. 86 μl of these were each mixed with 10 μl 10 × restriction buffer and 40 U restriction endonuclease and incubated for 4 to 8 h at the temperature recommended for the restriction enzyme. The DNA was then precipitated with ethanol in a manner known to the person skilled in the art and then taken up in 20 μl of water. The samples were then separated on an agarose gel according to methods known to the person skilled in the art, the DNA in the gel was denatured and transferred to the nylon membrane by means of capillary blot.

A corresponding probe for radioactive hybridization was created using the HighPrime (Röche) system. The membrane was first washed for 1 h at 65 ° C. with HybPuffer (1% (w / v) bovine sums albumin; 7% (w / v) SDS; 1 mM EDTA; 0.5 M sodium phosphate buffer, pH 7.2). prehybridized. The heat-denatured probe was then added and hybridization was carried out at 65 ° C. overnight. The blots were then washed as follows: rinsing once with 2xSSPE / 0.1% SDS; for 15 min. wash at 65 ° C with 2xSSPE / 0.1% SDS; for 15 min. wash at 65 ° C with lxSSPE / 0, 1% SDS and if necessary the last step was repeated again (20x SSPE is 3 M NaCl; 0.2 M NaH ₂ P0 ₄ ; 0.5 M EDTA; pH 7.4 ).

The hybridization was then analyzed using a phospho-imager (Molecular Imager FX, BioRad).

For example, total DNA from various plants cut with PstI and regenerated after transformation with pCB199-3 was hybridized with the aadA gene as a radioactively labeled probe (793bp PstI \ NcoI fragment from pCBl99-3). It was found that the lines CB199NTH-4, -6 and -8 had actually incorporated the aadA gene into the DNA. Furthermore, total DNA of. Cut with EcoRI and Xhol

CB199NTH-4, -6 and -8 hybridized with a radioactively labeled probe (1082 bp Bspl20I / SacI fragment from pCB199-3), which with part of the 16S rDNA hybridized. While, as expected, a band of approximately 3100 bp was detected in the wild type (untransformed plant), a band at 1750 bp was predominantly detected in the transplastomic lines, which results from the insertion of the insertion sequence from pCB199-3 into the piastom (FIG. 9 ). The plants obtained can be understood as predominantly homotransplastoma.

Example 5: Cloning of recombinases and the DSBI enzyme I-Ppo I

5.1 Cloning the recombinase from the phage ΦC31 and nuclear

Transformation as PLS fusion The recombinase from the phage ΦC31 (DSM No. 49156, DSMZ-German Collection of Microorganisms and Cell Cultures GmbH, Braunschweig, Germany) was amplified by means of PCR. The following oligonucleotides were used for this:

p65: ATGGATCCATATGGCCATGGCACAAGGGGTTGTG (SEQ ID NO: 38)

p66: CCCGGGCTCGAGCTGCAGCTACGCCGCTACGTC (SEQ ID NO: 39)

The PCR fragment obtained was cloned in pGEM-Teasy and sequenced. The resulting vector was named pCBl09-28. The gene coding for the recombinase was then fused to sequences which code for the IPP transit peptide and functionally coupled to the 35S promoter and before the ocs terminator. In addition to the element described, a nos promoter - pat - nos terminator element was present as a selection marker, which allowed the selection of transgenic plants for phosphinotricin. The finished binary construct was named pCBl93-17. The sequence of the insert is described under SEQ ID NO: 40 and contains the following elements:

Position base pair 7 to 535: 35S promoter

Position base pair 563 to 733: sequence coding for

IPP transit peptide

Position base pair 749 to 2590: gene coding for the ΦC31

recombinase

Position base pair 2615 to 2807: ocs terminator

The plasmid pCB193-17 was transformed by means of agrobacterium-mediated transformation in tobacco (N. tabcum cv. Petit Havana) according to methods familiar to the person skilled in the art. Transgenic plants from this transformation were designated CB193NTH. Expression of the ΦC31 recombinase was detected using RealTime PCR.

5.2 Cloning the recombinase from the phage TP901-1 and nuclear transformation as a PLS fusion

The recombinase from the phage TP901-1 (was present as a prophage in a lysogenic strain of Lactococcus lactis (strain 901-1)) was amplified by means of PCR. The following oligonucleotides were used for this:

p71: TGGATCCATATGGCCATGGCTAAGAAAGTAG (SEQ ID NO: 43)

p72: CCTCGAGCTGCAGTTAAGCGAGTTGG (SEQ ID NO: 44)

The PCR fragment of 1486 bp obtained was cloned into pCR2.1-TOPO (Invitrogen) and sequenced. The resulting vector was named pCBl27-2 (two differences were found in the sequence compared to the published sequence, one of which leads to an amino acid exchange (VIA)). The gene coding for the recombinase was then fused to sequences which code for the IPP transit peptide and functionally coupled to the 35S promoter and before the ocs terminator. In addition to the element described, a nos promoter - pat - nos terminator element was present as a selection marker, which allowed the selection of transgenic plants for phosphinotricin. The finished binary construct was named pCB245-37. The sequence of the insert is described under SEQ ID NO: 45 and contains the following elements:

Position base pair 12 to 540: 35S promoter

Position base pair 568 to 738: sequence coding for IPP transit peptide

Position base pair 748 to 2205: gene coding for the TP901-1

recombinase

Position base pair 2231 to 2422: ocs terminator

The plasmid pCB245-37 was transformed by means of agrobacterium-mediated transformation in tobacco (N. tabcu cv. Petit Havana) according to methods familiar to the person skilled in the art. Transgenic plants from this transformation were designated CB245NTH. The Expression of the TP901-1 recombinase was detected using RealTime PCR.

5.3 Cloning the Homing Endonuclease I-Ppol

The homing endonuclease I-Ppol was generated from 26 synthetically generated oligonucleotides by means of PCR based on the method of Stemmer WPC et al. (1995) Gene 164: 49-53 (SEQ ID NO: 4). The underlying sequence was derived from the published sequence (Accession No. M38131 nucleotides 86 to 577). Some mutations were introduced to remove restriction endonuclease recognition sites from the gene, but none of them required an altered amino acid sequence.

Example 6: Crossing master plants CB199NTH with plants expressing recombinase

6.1 Crossing master plants CB199NTH with ΦC31 recombinase expressing plants CB193NTH

Pollen from plants CB193NTH was dusted on the stigma of flowers from plants CB199NTH (lines -4, -6 and -8). The anthers of the corresponding flowers had previously been removed from the CB199NTH plants. The plants were kept in until maturity

Leave the greenhouse. The harvested seeds were laid out in vitro on cultivation medium (cf. Example 4; plus 500 mg / L spectinomycin, 10 mg / L phosphinotricin) and resistant plants were further grown under sterile conditions. These plants were designated CB199NTHxCBl93NTH. Explants from 4 to 8 week old plants were used for further bombardment experiments (see below).

6.2 Crossbreeding master plants CB199NTH with plants CB245NTH expressing TP901-1 recombinase

Pollen from plants CB245NTH was dusted on the stigma of flowers from plants CB199NTH (lines -4, -6 and -8). The anthers of the corresponding flowers had previously been removed from the CB199NTH plants. The plants were left in the greenhouse until the seeds ripened. The harvested seeds were laid out in vitro on cultivation medium (cf. Example 4; plus 500 mg / L spectinomycin, 10 mg / L phosphinotricin) and resistant plants were further grown under sterile conditions. These plants were designated CBl99NTHxCB245NTH. Explants from 4 to 8 week old plants were used for further bombardment experiments (see below). Example 7: Creation of a vector which is suitable for integration with the aid of recombinases into the piastome of corresponding master plants

The various elements necessary for such a vector were successively cloned into the backbone of pBLUESCRIPT. The specified sequence replaces the multiple cloning site of pBLUESCRIPT from Kpnl to Sacl, the Kpnl interface no longer being functionally present in the specified sequence due to the cloning strategy. The backbone of the vector corresponds to that of the pBLUESCRIPT. The vector obtained was designated pCB249-24. The insert is described by SEQ ID NO: 48 and comprises the following elements:

Elements :

Position base pair 861 to 910: attP recognition region for the

Recombinase from ΦC31

Position base pair 925 to 1031: Prpsl6 promoter from tobacco

Position base pair 1040 to 1060: 5'rbcL (5 'untranslated

Area of the rbcL gene from tobacco)

Position base pair 1076 to 1879: nptll gene

Position base pair 1916 to 2053: 3'rbcL (3 'untranslated

Area of the rbcL gene from tobacco)

Example 9: Creation of a vector which is suitable for integration with the aid of recombinases into the piastome of corresponding master plants and which enables a cut by a DSBI enzyme

The various elements necessary for such a vector were successively cloned into the backbone of pBLUESCRIPT. The specified sequence replaces the multiple cloning site of pBLUESCRIPT from Kpnl to Sacl, the Kpnl interface no longer being functionally present in the specified sequence due to the cloning site. The backbone of the vector corresponds to that of the pBLUESCRIPT. The vector obtained was designated pCB384-40. The insert is described by SEQ ID NO: 49 and comprises the following elements: i) Position base pair 70 to 765: homologous region to piastome sequences in the master plants CB199NTH - after integration of the insertion sequence under consideration here using PhiC31 recombinase, this region is duplicated as a direct "repeat" (same orientation) to the homologous sequences from CB199NTH. This allows the intrachromosomal recombination after the insertion of the insertion sequence, which leads to the elimination of the I-Ppol recognition region from the piastom copies already modified by insertion. It contains complementary from base pair 70 to

163 a 3 'psbA sequence and complementary from base pair 175 to 765 a portion of the aadA gene.

ii) Position base pair 861 to 910: attP recognition region for the recombinase from PhiC31

iii) Position base pair 925 to 1031: Prpsl6 promoter from tobacco

iv) position base pair 1040 to 1060: 5'rbcL (5 'untranslated region of the rbcL gene from tobacco)

v) Position base pair 1076 to 1879: nptll gene

vi) position base pair 1897 to 1974: 5 'psbA (5' untranslated region of the psbA gene from tobacco)

vii) position base pair 1975 to 2466: coding region for the homing endonuclease I-Ppol

viii) position base pair 2501 to 2638: 3'rbcL (3 'untranslated region of the rbcL gene from tobacco)

Example 10: Transformation of CBl99NTHxCBl93NTH plants with PCB249-24

The insertion sequence was first cut out of the plasmid in vitro and then circularized. 5 μg of the DNA from pCB249-24 were incubated for 2 hours with 10 U Sall (Röche) and 10U Xhol (Röche) in a 20 μl restriction mixture under the conditions specified by the manufacturer. The mixture was then incubated at 65 ° C. for 20 minutes. A ligation reaction was carried out at 14 ° C. overnight (addition of 5 μl lOx ligase buffer, 2U T4 ligase (Röche) to the restriction mixture and make up to 50 μl with water). Recircularization of the fragments was predominantly observed under the chosen conditions (not shown). The DNA was extracted from the ligation mixture using the Wizard Plus Minipreps DNA Purification System (Promega). cleaned. The ligation approach was used instead of bacteria. In contrast to the manufacturer's protocol, the DNA was eluted in 30 μl (instead of 50 μl) of water. The DNA purified in this way was used in a coating mixture as described in Example 4 (less than the 55 μl of water specified above was then used).

The DNA was then shot onto tobacco (Nicotiana tabacum cv. Petit Havana CB199NTH X CB193NTH) leaves as described in Example 4. In contrast to the information in Example 4, the selection is made for 30, 50 or 100 mg / L kanamycin.

Example 12: Transformation of CB199NTHxCBl93NTH plants with PCB384-40

The insertion sequence was first cut out of the plasmid in vitro and then circularized. 5 μg of the DNA from pCB384-40 were incubated with 10 U Sall (Röche) and 10U Xhol (Röche) in a 20 μl restriction mixture under the conditions specified by the manufacturer for 2 h. The mixture was then incubated at 65 ° C. for 20 minutes. A ligation reaction was carried out at 14 ° C. overnight (addition of 5 μl lOx ligase buffer, 2U T4 ligase (Röche) to the restriction mixture and make up to 50 μl with water). Recircularization of the fragments was predominantly observed under the chosen conditions

(Not shown) . The DNA was purified from the ligation mixture using the Wizard Plus Minipreps DNA Purification System (Promega). The ligation approach was used instead of bacteria. In contrast to the manufacturer's protocol, the DNA was eluted in 30 μl (instead of 50 μl) of water. The DNA purified in this way was used in a coating mixture as described in Example 4 (less than the 55 μl of water specified above was then used). The DNA was then shot onto tobacco (Nicotiana tabacum cv. Petit Havana CB199NTH X CB193NTH) leaves as described in Example 4. In contrast to the information in Example 4, the selection is made for 30, 50 or 100 mg / L kanamycin.

Example 13: Creating another vector (pCB456-2) for the introduction of a non-naturally occurring one

Detection region for the homing endonuclease I-Ppol and RE sequences (attB) for the recombinases φC31 and TP901-1 in the piastome of tobacco

The aim of this approach (analogously to example 3) was to create a further vector for plastid transformation which did not have any comprehensive homologies to sequences in the plastid genome owns. The selection marker aadA in this plasmid is under the control of a synthetic promoter, which is derived from the consensus sequence for E. coli σ70 promoters. A region downstream of the 3'psbA-1 gene from Synechocystis was used as the 3 'end. In contrast to the previously described vector pCB199-3 the RE sequence was located immediately downstream, but upstream introduced 3 psbA l ^λ sequence from Synechocystis in the molecule of the aadA gene which here. This allows you to create an operon after inserting an insertion sequence using a recombinase. The genes on the insertion sequence can then - optionally - be inserted without a promoter on the insertion sequence. After the insertion, corresponding genes in the insertion sequence then also come under the control of the synthetic promoter upstream of the aadA gene in the master plant. This allows - optionally - to create an operon structure consisting of the aadA and the genes introduced below in the piastom.

To create the plasmid pCB456-2, various elements were successively cloned into the base vector pCB42-94:

a) Synthetic promoter (complementary bp 1226-1260) b) Ribosome binding site (complementary bp 1214-1218) c) aadA gene (complementary bp 414-1208) d) Core recognition region for the homing endonuclease I-Ppol (complementary bp 331-345 ) e) attB RE sequence for the recombinases φC31 (bp 285-324) and TP901-1 (complementary bp 276-176) f) 3'psbA-l from Synechocystis (complementary bp 19-155)

The vector obtained in this way also conferred spectinomycin resistance in E. coli. Instead of the multiple cloning site in the base vector pCB42-94 (SEQ ID NO: 1) for the plastid transformation, this vector called pCB456-2 contains the elements listed within the framework of the nucleic acid sequence with SEQ ID NO: 50. Here too, the entire MCS is replacing Sequence (from Sacl to Kpnl) given.

Example 14: Creation of predominantly homotransplastomic master plants which contain a non-natural DSB recognition sequence and non-natural RE sequences (attB)

Analogously to pCB199-3 in Example 4, the vector pCB456-2 was introduced into the plastids of tobacco. Deviating from the description in Example 4, however, the shoots obtained were grown on growth medium which contained 30 g / L sucrose (instead of the game 4 specified 10 g / L), cultivated. The resulting plants were named CB456NTH. Southern hybridization identified 2 lines (CB456NTH-1 and -15) from the spectinomycin-resistant plants which were obtained after the transformation and which had inserted the insertion sequence from pCB456-2 into their piastome. In the Southern experiment, a probe was used (SEQ ID NO: 58) which was directed against a fragment of the 16SrDNA. This probe was suitable for detecting a fragment of approximately 3.1 kb from DNA which had been digested with EcoRI and which corresponds to the wild type. In contrast, an approximately 1.7 kb fragment was detected when the insertion sequence from pCB456-2 had been inserted into the corresponding piastom copies (see FIG. 11).

Example 15: Creating a sequence of the TP901-1 recombinase with a modified amino acid sequence

A PCR reaction was carried out with pCBl27-2 as the template and the oligonucleotides p354 (SEQ ID NO: 51) and p355 (SEQ ID NO: 52) as the primer, in order to insert a ribosome binding site upstream of the sequence coding for the TP901-1 recombinase and correct the described deviation in the amino acid sequence (see Example 5.2). The amplified 5 ^V region of the recombinase was cloned into pCR4Blunt TOPO (Invitrogen) and then ligated as a Spel \ HindiII fragment into the vector pCB127-2 treated with Xbal and HindIII. The resulting vector was named pCB487-20. The insert in this vector is described by SEQ ID NO: 60. It includes a ribosome site and codes for the TP901-1 recombinase (with a native amino acid sequence; SEQ ID NO: 61).

p354: 5 '-CGGATCCGGAGGCATCATGACTAAGAAAGTAGCAATCTATACACG-3' (SEQ ID NO: 51)

p355: 5 '-ATAAGCTTTATTCTCGATATCG-3' (SEQ ID NO: 52)

Example 16: Creation of a vector which is suitable for integration with the aid of recombinases into the piastome of master plants CB456NTH and additionally enables a cut by a DSBI enzyme

The various elements required for such a vector were successively cloned into the backbone of pBLUESCRIPT. The sequence given replaces the multiple cloning site of pBLUESCRIPT from Kpnl to Sacl. The backbone of the vector corresponds to that of the pBLUESCRIPT. The vector obtained was designated pCB557-l. The insert is through

SEQ ID No: 53 and includes the following elements:

a) Position base pair 150-460: homologous region to plastome sequences of the master plant CB456NTH - after integration of the insertion sequence considered here using mittelsC31 or TP901 recombinase, this region is duplicated as a direct “repeat” (same orientation) to the homologous sequences from CB456NTH This allows intrachromosomal recombination after insertion of the insertion sequence, which leads to the elimination of the I-Ppol recognition region from the plastom copies of the master plant which have already been modified by insertion. In this region homologous to sequences contained in CB456NTH are from base pair 150-244 (complementary ) Parts of the open reading frame 131 (0RF131) and base pair 337-431 the trnV gene from the tobacco piastome of the master plants. B) attP region of the TP901 recombinase (bp 7-85) c) attP region of the φC3l recombinase (complementary bp 96-145) d) Synthetic promoter (bp 462-495) e) Ribosome binding site (bp 514-518) f) nptll gene (bp 523-1326) g) ribosomes binding site (bp 1333-1337) h) I-Ppol coding gene (bp 1342-1833) Example 17 Creation of a vector which is suitable for integration with the aid of recombinases into the piastome of master plants CB456NTH and does not require an additional section through a DSBI Enzyme enables

For comparison to pCB557-l, a vector was created, the

Insertion sequence no longer encoded for a functional DSBI enzyme. This vector was created as follows. Vector pCB557-l was treated with the restriction enzyme Eco0109l. This removed a 230 bp fragment from the gene coding for the homing endonuclease I-Ppol. The remaining vector portion was recircularized in a ligation reaction. The resulting vector was named pCB559-6.

Example 15 Creation of a vector for the transient expression of the φC31 recombinase in plastids

In order to be able to transiently express the φC31 recombinase in plastids, it was cloned in the vector pCB554-3 downstream of a synthetic promoter. The following functional elements are present in the vector pCB554-3. a) Synthetic promoter (bp 28-62) b) Ribosome binding site (bp 70-74) c) Gene coding for the ΦC31 recombinase (bp 79-1920)

The insert in pCB554-3 is described by SEQ ID NO: 54. The vector backbone corresponds to that of the pBLUESCRIPT.

Example 16: Creation of an in vitro PCR product for the transient, plastid expression of the TP901 recombinase

The gene coding for the TP901-1 recombinase from pCB487-20 was fused with a synthetic promoter in a suitable orientation. The fusion product is then amplified by means of PCR; the resulting product is called PCR TP-Rek. The primers p268 to be used in the PCR

(SEQ ID NO: 55) and p72 (SEQ ID NO: 56) have the sequence:

p268: TCGACTTGACATTCACTCTTCAATTATCTATAATGATACATG (SEQ ID NO: 55) p72: CCTCGAGCTGCAGTTAAGCGAGTTGG (SEQ ID NO: 56)

The resulting PCR product (TP-Rek) comprises 1527 bp (SEQ ID NO: 57) and comprises the following elements:

a) Synthetic promoter (bp 6-40) b) Ribosome binding site (bp 48-52) c) Gene coding for the TP901 recombinase (bp 57-1514)

Example 17: Transformations of CB456NTH master plants for the integration of an insertion sequence by means of recognition site-specific recombination

In these experiments, the insertion sequence or the entire insertion construct is to be incorporated into the plastome sequence of the master plants CB456NTH by the transient expression of the recombinase in plastids. The corresponding DNA is applied to the gold particles analogously to Example 4. The following combinations of DNA are introduced simultaneously into the plastids of the master plants CB456NTH. If only the insertion sequence is introduced into the plastids, this is first cut out of the insertion construct in vitro as described in Example 12 (in the case of pCB559-6 and pCB557-l this is done with the help of the restriction endonuclease Kpnl) and circularized. If the entire, untreated insertion construct is used in the transformation, the vector backbone is also incorporated into the plastids.

The particle cannon transformation process is essentially as described in Example 14 for the creation of CB456NTH. As a major difference from the above description, the bombardment medium is replaced by regeneration medium and instead of spectomycin, kanamycin is used as a selective agent. In brief: The bombarded explants remain on regeneration medium for 2 days and are then cut into segments of approximately 0.5 cm2. These are incubated for 4 to 6 weeks on regeneration medium plus 30 mg / L kanamycin. The explants are then incubated for 7 to 10 weeks on regeneration medium plus 50 mg / L kanamycin. The explants are then - optionally - incubated on regeneration medium plus 80 mg / L kanamycin. Emerging plants are again placed on growth medium (plus kanamycin) until they form roots. Rooting can also take place in the absence of the selective agent.

Claims

claims

1. A method for the sequence-specific integration of a DNA sequence into the plastid DNA of a plant organism or cells derived therefrom, characterized in that

a) the plastid DNA molecules of said plant organism or cell derived therefrom contain at least one recombinase recognition sequence R1 and

b) i) at least one transformation construct comprising an insertion sequence flanked at least on one side by at least one further recombination sequence

R2 as well

ii) at least one sequence-specific recombinase suitable for inducing recombination events on said recombination sequences R1 and R2

in at least one plastid of said plant organism or cells derived therefrom, and

c) the insertion sequence is inserted into the plastidic DNA with recombination of R1 with R2, and

d) plants or cells are isolated in which the insertion sequence has been inserted into the plastid DNA molecules.

2. The method according to claim 1, characterized in that the functionality of the recombination sequences resulting from recombination of R1 with R2 is reduced so that the insertion of the insertion sequence is preferred to the excision thereof by at least a factor of ten.

Sign.

3. The method according to any one of claims 1 or 2, characterized in that the transformation construct contains at least one of the following elements

i) Expression cassette for a recombinase

ii) Positive selection markers

iii) negative selection markers

iv) reporter genes

v) Origins of replication

vi) Multiple cloning regions

vii) "Border" sequences for Agrobacterium transfection

viii) sequences which enable homologous recombination or insertion into the genome of a host organism

ix) Expression cassette for an enzyme, suitable for the induction of DNA double-strand breaks on the recognition sequence for the targeted induction of

DNA Doppe1strangbrüchen

4. The method according to any one of claims 1 to 3, characterized in that the recombinase is selected from the group of recombinases consisting of Cre, Flp, ΦC31 recombinase, TP901-1 recombinase, gin, Cin, pin, λ recombinase, R4 integrase and Hin Invertase.

5. The method according to any one of claims 1 to 4, characterized in that the recombinase as a fusion protein with a

Plastid localization sequence is expressed.

6. The method according to any one of claims 1 to 5, characterized in that the recombinase is encoded by a sequence according to SEQ ID NO: 41, 42, 46 or 47.

7. Expression cassette containing a nucleic acid sequence coding for a recombinase selected from the group of resolvases / invertases under the control of a promoter active in plant plastids.

8. Expression cassette containing a nucleic acid sequence coding for fusion protein from a plastid localization sequence and a recombinase selected from the group of resolvases / invertases under control

5 a promoter active in the plant cell nucleus.

9. Expression cassette according to one of claims 7 or 8, characterized in that the recombinase is selected from the group consisting of ΦC31 recombinase, R4

10 recombinase, Hin recombinase, TP901-1 recombinase, gin, cin and pin.

10. Plant organism obtained by a method according to any one of claims 1 to 6.

15

11. Plant organism containing an expression cassette according to one of claims 7 to 9.

12. Plant organism according to one of claims 10 or 11 20 selected from the group consisting of Arabidopsis thaliana,

Tobacco, tagetes, wheat, rye, barley, oats, rapeseed, maize, potatoes, sugar beet, soy, sunflower, pumpkin or peanut.

25 13. Cell cultures, organs, tissues, parts or transgenic reproductive material derived from a plant organism according to claims 10 to 12.

14. Use of an organism according to any one of claims 10 to 30 12 or cell cultures, organs, tissues, parts or transgenic reproductive material derived therefrom according to claim 13 as food, feed or seed or for the production of pharmaceuticals or fine chemicals.

35

40

45