EP1042478A1

EP1042478A1 - Herbicide binding proteins and transgenic plants containing them

Info

Publication number: EP1042478A1
Application number: EP98960019A
Authority: EP
Inventors: David Charles Holt; Paul Glyn Jones
Original assignee: Zeneca Ltd
Current assignee: Syngenta Ltd
Priority date: 1997-12-19
Filing date: 1998-12-15
Publication date: 2000-10-11
Also published as: GB9726955D0; WO1999032630A1; AU1570699A

Abstract

A protein which is capable of binding a herbicide, said protein comprising the sequence depicted as SEQ ID No 2 or an active part thereof or SEQ ID No 4 or an active part thereof with the proviso that the active part is not that depicted as SEQ ID Nos 35 to 38.

Description

HERBICIDE BINDING PROTEINS AND TRANSGENIC PLANTS CONTAINING THEM

The present invention relates to recombinant DNA technology, and in particular to the production of transgenic plants which exhibit substantial resistance or substantial tolerance to herbicides when compared with non transgenic like plants. The invention also relates, inter alia, to the nucleotide sequences (and expression products thereof) which are used in the production of, or are produced by, the said transgenic plants.

Plants which are substantially "tolerant" to a herbicide when they are subjected to it provide a dose/response curve which is shifted to the right when compared with that provided by similarly subjected non tolerant like plants. Such dose/response curves have "dose" plotted on the x-axis and "percentage kill", "herbicidal effect" etc. plotted on the y- axis. Tolerant plants will typically require at least twice as much herbicide as non tolerant like plants in order to produce a given herbicidal effect. Plants which are substantially "resistant" to the herbicide exhibit few, if any, necrotic, lyric, chlorotic or other lesions when subjected to the herbicide at concentrations and rates which are typically employed by the agrochemical community to kill weeds in the field. Hereinafter the words (i) "tolerant" and (ii) "resistant" when used individually mean "tolerant and/or resistant".

It is particularly preferred that the plants are substantially resistant or substantially tolerant to herbicides which inhibit photosynthesis, of which paraquat and diquat (and structurally related analogues) are notable examples.

There are a number of methods currently used to produce herbicide resistant plants. These include for example, the introduction into a plant of a target site for a herbicide which has been modified such that it is no longer sensitive to the herbicide. Herbicide resistance may also be produced by over-expressing genes coding for the herbicide target protein such that more herbicide is required to suppress and inactivate the target protein. The over- expression of genes coding for proteins which are produced in response to the presence of herbicide and which act to neutralise its effect may also be used to produce herbicide resistance. An example of this approach is where the presence of herbicide results in production of free radicals in the plants, the overproduction in the plant of proteins which mop up the radicals may result in herbicide resistance. Transformation of plants with genes coding for proteins which metabolise the herbicide such as herbicide degrading enzymes is a yet further means of producing herbicide resistant plants.

The present invention provides, ter alia, a method of producing plants which are tolerant or resistant to herbicides. This approach involves the use of herbicide binding proteins advantageously to sequester the herbicide, for example at the cell surface or in the vacuoles of a treated plant. Sequestration at the cell surface prevents the entry of the herbicide into the cell so that the herbicide cannot reach its intracellular target and exert any significant cytotoxic effect. Similarly, sequestration in the vacuole effectively removes the herbicide from its target site.

The invention offers the further advantage of inhibiting the mobility of the herbicide from the application site to the whole plant therefore preventing the herbicide reaching particularly sensitive organs.

Strategies of producing herbicide resistance are often based on a detailed knowledge of the mechanism and site of action of herbicides. It is, however, possible according to the invention, to produce herbicide tolerant plants without knowledge of the mode of action of the herbicide.

A further advantage of this invention is that tolerant plants can be produced against herbicides which have more than one target site. Central to solving the problem which the present invention addresses, viz. the production of herbicide resistant or tolerant plants, is the provision of novel nucleotide sequences which can be used in the transformation of plant material.

According to the present invention there is provided a protein which is capable of binding a herbicide, said protein comprising the sequence depicted as SEQ ID No 2 or an active part thereof or SEQ ID No 4 or an active part thereof with the proviso that the active part is not that depicted as SEQ ID Nos. 35 to 38. It is particularly preferred that the herbicide is paraquat or diquat or a structurally related analogue thereof. A structural analogue is defined as a second compound having a similar activity to a first and which is capable of reacting with antibodies raised to the first compound to which it possesses a similar three-dimensional structure - at least in part. As used herein the term "protein which is capable of binding a herbicide" is used to denote a protein which does not catalyse a chemical alteration in the herbicide upon binding. Examples of proteins which are capable of binding herbicides include antibodies e.g. a monoclonal antibody having full length heavy and light chains or a fragment thereof such as a Fab, Fab' '(Fab' )₂ or Fv fragment: a single chain antibody fragment e.g. a scFv; a light chain or heavy chain monomer or dimer; or multivalent mono-specific antigen binding proteins comprising two, three, four or more antibodies or fragments thereof bound to each other by a connecting structure; receptor proteins or enzymes where the catalytic domain has been inactivated e.g. by mutagenesis while maintaining their binding capability; DNA/protein or carbohydrate/cellulose binding proteins where the binding properties have been altered to allow herbicide binding, and proteins having herbicide binding domains, as revealed by known phage display techniques. The present invention further provides a protein which is capable of binding a herbicide, preferably paraquat or diquat or a structurally related analogue thereof, said protein comprising the amino acid sequence depicted as SEQ ID No 2 or an active part thereof or SEQ ID No 4 or an active part thereof wherein the said active part has a K_a greater than 1.24 x 10⁶ liters/mole for paraquat. The affinity constant K_a can be expressed as:

K_a = [AgAb]/[Ag][Ab]

Where Ag denotes a single antigenic determinant and Ab denotes a single antigen-binding site and the square brackets indicate the concentration of each component at equilibrium. Thus the affinity constant can be determined by measuring the free Ag required to fill half of the antigen-binding sites on the antibody. When half the sites are filled, [AgAb] = [Ab] and the K_a = l/[Ag]. Thus the reciprocal of the antigen concentration that produces half-maximal binding is equal to the affinity constant of the antibody for the antigen.

Antibodies and fragments thereof for use in the invention will have a K_a in the range of about 1.2 x 10⁶ to 10^I2l/mole. It is preferred that the K, is around 1.0 x 10⁷ more preferably 1.0 x 10⁸ , more preferably 1.0 x 10⁹, more preferably 1.0 x 10¹⁰ and even more preferably 1.0 x 10¹². Generally according to the method of the invention the antibody will be present in the transgenic plant in equimolar amounts with respect to the herbicide present at the site of sequestration in the plant. The binding affinity of the binding protein for use in the method of the invention may be measured using techniques well known in the art.

The present invention further provides a protein which is capable of binding a herbicide, preferably paraquat or diquat or a structurally related analogue thereof, said protein comprising the sequence depicted as SEQ ID No 4 or an active variant thereof comprising the region shown in SEQ ID No 4 at amino acid number 35 to 45 or amino acid number 25 to 35 or amino acid number 26 to 36 or amino acid number 27 to 37 or amino acid number 28 to 38 or amino acid number 29 to 39 or amino acid number 30 to 40. or amino acid number 31 to 41 or amino acid number 32 to 42 or amino acid number 33 to 43 or amino acid number 34 to 44 or any conservative substitution of any amino acid within said active variant providing that the resulting K_a of the protein or variant for paraquat is greater than 1.24 x 10⁶ liters/mole.

The present invention further provides a protein which is capable of binding a herbicide, preferably paraquat or diquat or a structurally related analogue thereof, said protein comprising the sequence depicted as SEQ ID No 2 or an active part thereof which is linked to the sequence depicted as SEQ ID No 4 or an active part thereof, said sequences optionally comprising cysteine residues disposed in such a way as to facilitate the formation of disulphide cross bridges between them. It is particularly preferred that the said sequences are linked via a synthetic linker such as the one depicted as SEQ ID No. 27. It is further preferred that the said proteins are spaced apart by a linker which may comprise at least three consecutive cysteine residues, or be the hinge region of an IgG3 type antibody.

The present invention further provides a polynucleotide sequence encoding the protein or active part or variant described above. In a preferred embodiment the polynucleotide comprises the sequence depicted as SEQ ID No 1 and/or SEQ ID No 3. Also provided is a polynucleotide sequence which is the complement of one which binds to any polynucleotide described above, at a temperature of between 60°C and 65°C in 0.3 strength citrate buffered saline containing 0.1% SDS followed by rinsing at the same temperature with 0.3 strength citrate buffered saline containing 0.1% SDS and which still encodes a protein which is capable of binding a herbicide, preferably paraquat or diquat or a structurally related analogue thereof, with the proviso that the polynucleotide sequence does not encode a protein which has a K_a of less than or equal to 1.24 x 10⁶ liters/mole for paraquat.

The present invention further provides a DNA construct comprising in sequence a promoter region which is operable in plants operably joined to a polynucleotide sequence described above and a transcription termination region. The polynucleotide may be codon- optimised or otherwise altered, i.e. the protein and/or polypeptide encoding regions may be modified in that mRNA instability motifs and/or fortuitous splice regions may be removed, or plant preferred codons may be used so that expression of the thus modified polynucleotide in a plant yields substantially similar protein or polypeptide having a substantially similar activity/function to that obtained by expression of the unmodified polynucleotide in the organism in which the protein encoding regions of the unmodified polynucleotide are endogenous.

The DNA construct may further comprise (a) transcriptional enhancing elements; and/or (b) regions encoding non translated translational enhancing sequences, preferably Omega and Omega prime; and/or (c) regions encoding non translated sequences such as intron sequences; and/or (d) regions encoding target sequences which are capable of directing transcription products to either intracellular organelles, intracellular compartments, cell membranes or to the outside of the cell. It is preferred that the regions encoding the said target sequences are selected from the group depicted as SEQ ID Nos. 40 to 47. B It is further preferred that the said regions encoding the target sequences encode any one of the sequences depicted as SEQ ID Nos. 28 to 30. Further examples of suitable signal or targeting sequences include human or murine derived sequences such as immunoglobulin signal sequences, e.g. IgG light chain signal sequences or IgG heavy chain signal sequences or signal sequences derived from alpha amylase from barley aleurone; AFP signal sequences e.g. the signal sequence from Rs-AFPl ( F R Terras et al Plant Cell (1995) 7 573-588); the signal sequence from germin e.g. from wheat ( B G Lane et al Journal Biol. Chem. (1991) 266 (no 16) 10461- 10469) or the patatin signal sequence from potato tuber which may be used for vacuolar targeting. Various other target sequences may be used to direct the proteins of the present invention to a particular location within the plant. It is preferred that the target sequence is capable of directing the herbicide binding protein to or through an intracellular membrane or extracellularly.

Various plant operable promoters may be used in the DNA constructs according to the present invention. It is particularly preferred that the promoter is constitutive and provides for high level expression of the sequence 3' of it. Such promoters are known and include the CaMV35S or FMV35S promoters, for example. Alternatively the promoter may be tissue (which term includes seed) specific. It is further preferred that the promoter is Maize flag leaf specific. The promoter may also be a controllable promoter which may be induced chemically, developmentally or hormonally. The promoter may, under certain circumstances, be switchable such as the alcA/alcR gene switch described in published International Patent Application No. WO93/21334; the GST promoter switch described in published International Patent Application Nos. WO90/08826 and WO93/01294 and the RMS switch system described in published International Patent Application No. WO90/08830, the teachings of which are incorporated herein by reference.

Down regulation of the promoter may be achieved by use of repression proteins, anti- sense, partial sense and operator/repressor systems such as the lac system.

Other promoters which are applicable to the constructs and methods of the present invention may be identified and/or obtained using methods which are well known, documented and used within the art.

Suitable transcription terminators for use in the constructs according to the present invention include, for example, the known CaMV35S, NOS, OCS and E9 terminators.

The DNA construct according to the present invention may additionally further comprise a polynucleotide sequence which is capable of imparting to a plant, substantial resistance or substantial tolerance to a further herbicide. Preferably the further herbicide is N- phosphonomethylglycine or an agriculturally acceptable salt or ester thereof.

The present invention also provides a plant transformation vector comprising a DNA construct as described above. The plant transformation vector may additionally include a known selectable or screenable marker gene such as those that provide for antibiotic resistance e.g. kanamycin resistance or those providing for herbicide resistance e.g. glufosinate resistance.

The invention further provides a method of producing plants which are capable of producing herbicide binding proteins comprising inserting into a plant cell a polynucleotide, construct or a plant transformation vector as described above, regenerating morphologically normal fertile plants or plant parts therefrom and selecting from the population of regenerants those morphologically normal fertile plants or plant parts which are capable of producing the said herbicide binding proteins. Known plant transformation techniques include, for example, Agrobacterium mediated transformation; electroporation; micro-injection, and the use of the micro-projectile gun or 'whiskers' silicon carbon fibres. The methods of the present invention are also applicable to further transform plant material which is already resistant to a herbicide, preferably N-phosphonomethylglycine or an agriculturally acceptable salt or ester thereof. The transformed material may then be regenerated into whole plants, by known means (including somatic embryogenesis), in which the new nuclear material is stably incorporated into the genome. Also provided are morphologically normal fertile plants and plant parts produced according to above mentioned method. It is preferred that the morphologically normal fertile plants are small grain cereals, oil seed crops, fibre plants, fruit, vegetables, plantation crops and trees. It is further preferred that the plants are selected from the group consisting of soybean, cotton, tobacco, sugarbeet, oilseed rape, canola, flax, sunflower, potato, tomato, alfalfa, lettuce, maize, wheat, sorghum, rye, barley, oat, turf grass, forage grass, sugar cane, pea, field bean, rice, pine, poplar, apple, grape, banana, citrus or nut plants and the progeny and seeds thereof. Plants produced according to the present invention may also be used in a breeding program with plants which are already resistant to a further herbicide to provide plants which are substantially resistant and/or substantially tolerant to multiple herbicides. It is preferred that the said further herbicide is N- phosphonomethylglycine or an agriculturally acceptable salt or ester thereof. In the method of the invention the transformed material will express the herbicide resistance conferring regions in an amount effective to improve the tolerance of the plant to said herbicide when compared with plants regenerated from non-transformed like material. Preferably the expression of the binding protein or fragment thereof in homozygous plants is at least 0.5% of total plant protein. The method of the invention may be used to improve the herbicide tolerance of plants which previously showed some low level tolerance and also those which previously showed no measurable tolerance to the herbicide.

The invention still further provides a method of selectively controlling weeds in a field said field comprising weeds and crop plants said method comprising contacting said field with paraquat, characterised in that the said crop plants are the plants produced according to the present invention or the progeny thereof. The invention still further includes a variation of the method disclosed in the immediately preceding paragraph, wherein a pesticidally effective amount of one or more of an insecticide, fungicide, bactericide, nematicide and anti-viral is applied to the field, i.e. to the crop plants and weeds either prior or after application to the field of the herbicide. The invention still further includes the use of the polynucleotide or vector of the invention in the production of morphologically normal fertile whole plants which are substantially resistant or tolerant to herbicides which inhibit photosynthesis by accepting electrons from photosystem I thus generating free radicals which cause lipid peroxidation, or (ii) by blocking electron transport in photosystem II- The herbicide may be paraquat and/or diquat, or structurally related analogues thereof-

Antibodies and antibody fragments to herbicides may be produced using methods well known in the art see for example, Kohler and Milstein (Nature 256 479-495 1975); Graham et al (J. Chem Tech. Biotechnol. 6 _ 279-289 (1995); Niewola et al (Clinica Chimica Acta 148 149-156 (1985)) and Ward et al (Protein Engin 6 981-8 (1993)) and this is described further in the examples herein.

The present invention further provides for the use of a polynucleotide, construct or a plant transformation vector as described above, in the production of morphologically normal fertile plants which are substantially resistant or tolerant to herbicides which inhibit photosynthesis by accepting electrons from photosystem I thus generating free radicals which cause lipid peroxidation, or (ii) by blocking electron transport in photosystem II. Preferably the herbicide is paraquat or diquat or a structurally related analogue thereof.

The invention further provides for the use of a polynucleotide comprising a protein encoding region having the sequence depicted as SEQ ID No. 1 and/or SEQ ID No. 3 as a selectable or screenable marker gene in the selection or screening of transgenic plant material.

The present invention further provides an open reading frame coding for a functional paraquat binding protein having at least 99% probability of being the same as SEQ ID No. 1 or SEQ ID No. 3 as recognised by the FASTA algorithm. FASTA version 3.0t82, November 1, 1997 as described in W.R. Pearson & D.J. Lipman. PNAS (1988) 85: pp2444-2448.

The present invention further provides an open reading frame coding for a functional paraquat binding protein having a smallest sum probability of 5 x 10^"45 of being the same as SEQ ID No. 1 or SEQ ID No. 3 as recognised by the BLASTP 2.0alOMP-WashU algorithm. This is described in Gish et al. (1990) Basic local alignment search tool. J. Mol. Biol. 215: pp403.

The present invention further provides an open reading frame coding for a variable heavy antibody domain which ' can be used to produce a functional paraquat binding protein being 76-100% identical to SEQ ID No. 4 according to the Needleman and Wunsh equation. This is described in Needleman & Wunsch. (1970) J. Mol. Biol. Vol.48. p443. In this specification the term "conservative substitution" with reference to an amino acid means any conservative replacement which does not affect the activity and/or function of the protein when compared with the unmodified protein. In particular conservative replacements may be made between the following groups viz.

(a) Alanine, Serine, Glycine and Threonine

(b) Glutamic acid and Aspartic acid

(c) Arginine and Lysine

(d) Isoleucine, Leucine, Valine and Methionine (e) Phenylalanine, Tyrosine and Tryptophan

The invention is further illustrated in the following non-limiting examples and with reference to the following Sequence Listings and drawings of which:

Figure 1 shows a diagrammatic map of plasmid pMJB 1.

Figure 2 shows a diagrammatic map of plasmid pJRIi.

Figure 3 shows a diagrammatic map of plasmid pDmAMPULC.

Figure 4 shows a diagrammatic map of plasmid pDm.

Figure 5 shows a diagrammatic map of plasmid pBinl9. Figure 6 shows a comparison of PQ-ZEN1 scAb binding against a paraquat BSA conjugate.

SEQ ID No. 1 shows a nucleotide sequence comprising a region encoding part (SEQ

ID No. 2) of a variable light chain of an antibody which binds paraquat.

SEQ ID No. 3 shows a nucleotide sequence comprising a region encoding part (SEQ ID No. 4) of a variable heavy chain of an antibody which binds paraquat.

SEQ ID Nos. 5, 6, and 9 -14 show PCR primers.

SEQ ID No. 7 and 8 show polylinkers.

SEQ ID Nos 15 and 16 show oligonucleotides which are used in the production of vector pNG3-Vkss and SEQ ID Nos. 17 and 18 show oligonucleotides which are used in the production of vector pNG4-VHss

SEQ ID No. 19 shows a sequence encoding the human light chain kappa constant region of an antibody. SEQ ID No. 20 depicts a sequence comprising the coding region (SEQ ID No. 24) for the human heavy chain IgG2CHl' constant region (SEQ ID No. 23) of an antibody.

SEQ ID No. 22 depicts a sequence encoding HUIgGlCHl" (SEQ ID No. 21).

SEQ ID No. 26 depicts a sequence encoding HUIgG2CHl' (SEQ ID No. 25). SEQ ID No. 27 depicts a synthetic amino acid linker.

SEQ ID No. 28 depicts the Dhalia signal peptide.

SEQ ID No. 29 depicts the Tomato PG (Polygalacturonase) signal peptide.

SEQ ID No. 30 depicts the Tobacco PRS signal peptide.

SEQ ID Nos. 31 to 34 depict PCR primers. SEQ ID Nos. 35 to 38 depict prior art sequences.

SEQ ID No. 39 depicts the nucleotide sequence from the EcoRl to Hind III restriction sites in pMJB 1.

SEQ ID Nos. 40 to 47 depict various targeting sequences.

Cloning and sequencing of the variable regions of POXBl/2 antibody heavy and light chain genes- Preparation of cytoplasmic RNA

There are several procedures for the isolation of poly A+ mRNA from eukaryotic cells (Sambrook J., Fritsch E.F., Maniatis T., Molecular Cloning; A Laboratory Manual, Cold Spring Harbor Laboratory Press, Second Edition, 1989, Chapter 8, p3 herein referred to as "Maniatis"). In this particular case cytoplasmic RNA is prepared as described by Favoloro et al., Methods in Enzymology 65, 718-749, from a frozen hybridoma cell pellet containing lxlO⁹ cells which are stored at -80°C.

The cells are resuspended in 5ml ice-cold lysis buffer (140mM NaCl, 1.5mM McCl₂, lOmM Tris-HCl pH 8.6 and 0.5% NP40 (a polyglycol ether nonionic detergent;

Nonyllphenoxy Polyethoxy Ethanol, Sigma Cat. No. 127087-87-0)) containing 400u of a ribonuclease inhibitor (RNAguard; Pharmacia Cat. No. 27-0815-01) and vortexed for 10s. This solution is overlayed on an equal volume of ice cold buffer containing 24% (w/v) sucrose and 1% NP-40 and stored on ice for 5 min. The preparation is then centrifuged at 4000 rpm for 30 min at 4 °C in a bench top centrifuge (Sorval RT6000B) after which, the upper cytoplasmic phase is removed to an equal volume of 2 x PK buffer (200mM Tris (pH7.5), 25mM EDTA, 300mM NaCl and 2% (w/v) SDS). Proteinase K (Sigma, Cat No. P2308) is added to a final concentration of 200 μg/ml and the mixture incubated at 37°C for 30 min.

The preparation is extracted with an equal volume of phenol/chloroform, the aqueous phase removed and 2.5 vol ethanol added and mixed. This solution is then stored at -20 °C overnight. RNA is collected by centrifugation (4000 φm, 30 min at 4 °C in a bench top centrifuge, Sorval RT6000B), the supernatant decanted and the pellet dried in a vacuum dessicator after which it is dissolved in 250 μl diethylpyrocarbonate (DEPC)-treated water (prepared as described in Maniatis, referenced above). The RNA content is measured by spectrophotometry and the concentration calculated assuming an absorbance at 260nm of 1=40 μg/ml.

Preparation of first strand variable region cDNA

A number of methods for the synthesis of cDNA are reviewed in Maniatis (Chapter 8). The oligonucleotide primers used are mainly based on those proposed by Marks et al. J. Mol. Biol (1991) 222, 581-597. The cDNA in this case is prepared as described below. RNA (5mg) is mixed in a microcentrifuge tube with 10 μl 5x reverse transcriptase buffer [250mM Tris (pH8.3), 40mM MgCl₂ and 50mM DTT], 1 μl forward primer (25 pM), 10 μl 1.25mM dNTPs, 5 μl lOmM DTT, 0.5 μl RNAguard (Pharmacia) to which DEPC-treated H₂O is added to obtain a volume of 50 μl. The reaction mix is heated to 70 °C for 10 min and then cooled slowly to 37 °C, after which 100 u (0.5 μl) M-MLV reverse transcriptase (Pharmacia Cat. No 27-0925-01) are added and the reaction incubated at 37 °C for 1 hour. The forward primer used for the generation of the light chain cDNA is oligonucleotide CK2FOR (SEQ ID NO: 1) which is designed to hybridise to the CK constant region of murine kappa light chain genes. For the heavy chain cDNA the forward primer CGI FOR (SEQ ID NO:2) was used which hybridises to the CHI constant domain of murine IgGl .

Isolation of antibody gene fragments bv PCR

Isolation of PQXB1/2 heavy and light chain variable region genes is performed using the cDNA prepared as described above as template. General reaction conditions are as follows. To 5 μl of the cDNA reaction is added 5 μl dNTPs (2.5 mM), 5 μl lOx Enzyme buffer (500mM Kcl, lOOmM Tris (pH8.3), 15mM MgCl₂ and 0.1% gelatin), 1 μl of 25pM/μl back primer, 1 μl of 25pM/μl forward primer, 0.5 μl AmpliTaq (a thermostable DNA polymerase, Perkin-Elmer Cetus) and DEPC-treated water to obtain a volume of 50 μl. The PCR conditions are set for 25 cycles at 94 °C for 90s; 55 °C for 60s; 72 °C for 120s, ending the last cycle with a further 72 °C for 10 min incubation.

Using the general reaction conditions, the forward primer used for the generation of the light chain cDNA is oligonucleotide CK2FOR, (SEQ ID NO: 5) and the for the heavy chain cDNA oligonucleotide CGIFOR (SEQ ID NO: 6). A variety of different back primers are synthesised for both heavy and light chains based on the DNA sequence (B M Graham et al J. Chem. Tech. Biotechnol (1995) 63 279- 89). A number of reactions are performed using these back primers and those that produced the desired PCR product are used for obtaining the variable regions of the heavy and light chains. Reaction products are analysed on a 2% agarose gel. Products of the expected size, are excised, the DNA purified using GeneCLEAN (Stratech Scientific, following the manufacturers instructions - see below) and the DNA fragment digested with the relevant restriction enzymes in preparation for subsequent vector cloning. The GENECLEAN kit contains 1) 6M sodium iodide 2) a concentrated solution of sodium chloride, Tris and EDTA for making a sodium chloride/ethanol/water wash; 3) Glassmilk (TM)- a 1.5 ml vial containing 1.25 ml of a suspension of a specially formulated silica matrix in water. This is a technique for DNA purification based on the method of Vogelstein and Gillespie published in Proceedings of the National Academy of Sciences USA (1979) Vol 76, p 615. Alternatively any of the methods described in "Molecular Cloning - a laboratory manual" Second Edition, Sambrook, Fritsch and Maniatis (Cold Spring Harbor Laboratory, 1989) can be used. Briefly, the Geneclean procedure is as follows. To 1 volume of gel slice is added 3 volumes of sodium iodide solution from the kit. The agarose is melted by heating the mix at 55 °C for 10 min then Glassmilk (5-10ml) is added, mixed well and left to stand for 10 min at ambient temperature. The glassmilk is spun down and washed 3 times with NEW WASH (500ml) from the kit. The wash buffer is removed from the Glassmilk which is allowed to dry in air. The DNA is eluted by incubating the dried Glassmilk with water (5- 10ml) at 55 °C for 5-10 min. The aqueous supernatant containing the eluted DNA is recovered by centrifugation. The aqueous supernatant containing the eluted DNA is recovered by centrifugation. The elution step can be repeated and supernatants pooled.

Isolation of Antibody Gene Fragments using Overlapping Oligonucleotides

The sequences for the variable regions of the light and heavy chains of PQXB1/2 (B M Graham et al J. Chem. Tech. Biotechnol. (1995) 63 279-89) can also be synthesised by a variety of methods including those described by Edwards (1987) Am. Biotech. Lab 5, 38-44, Jayaraman et al (1991) Proc. Natl. Acad. Sci. 88, 4084-4088, Foguet and Lubbert (1992) Biotechniques 3, 674-675 and Pierce (1994) Biotechniques \6_, 708. Preferably the sequences for the variable regions of the heavy and light chains are prepared by a PCR method similar to that described by Jayeranan et al (1991) Proc. Natl. Acad. Sci. USA 88. 4084-4088.

Oligonucleotide primers can be designed to alter codon usage as necessary to eliminate unwanted internal restriction sites i.e.EcORI, Hind III, NcOI and Sma 1.

Cloning of the PCR products into Bluescript KS+ vector

For each antibody fragment, both the 5' region (back primer) oligonucleotide and the 3 ' region (forward primers) can introduce a restriction site. The discrete PCR products are for both the VH and VK PCR reactions and are therefore able to be cloned into the Bluescript vector KS+ (Stratagene Cloning Systems) via the appropriate enzyme restriction sites using standard DNA manipulation methods. DNA is prepared from the clones obtained and rigorous sequencing of several clones of each construct performed using automated fluorescent sequencing equipment (Applied Biosystems). The sequences are reviewed, compared and aligned using suitable computer software. A clone containing the light chain is designated VKPQ1 and a clone containing the heavy chain sequence is designated VHPQ1.

Construction of chimaeric light chain and heavy chain fd' genes The heavy and light chain genes which are cloned into Bluescript VKPQ1 and

VHPQ1 are isolated by PCR using primers which allow specific amplification of only the variable region of the appropriate genes but also introduced new unique enzyme restriction sites. These restriction sites enable the variable region gene fragments to be cloned in frame with DNA fragments coding for both the appropriate antibody signal sequences and human constant regions. The signal and constant region sequences for the light and heavy chain Fd' have each been previously cloned into pNG3 and pNG4, derivatives of the pSG5 Eukaryotic plasmid expression vector.

The vector pNG3 was prepared as follows. Plasmid pSG5 (Stratagene, Cat No. 216201) was digested with Sail and Xbal to remove the existing SV40 promoter and polylinker sequence. A new polylinker was introduced by use of oligonucleotides SEQ NOS 7 and 8 which were hybridised and cloned into the Sail and Xbal cut pSG5 plasmid to give plasmidpNG 1. The pNG 1 plasmid was cut with Bglll and Hindlll and the Bglll-Hindlll CMV promoter fragment from pcDNA3 (Invitrogen, Cat. No V790-20) cloned into this site to give plasmid pNG2. Finally, the polyA region from pSG5 was isolated by PCR as described above but using oligonucleotide sequences SEQ ID NOS: 9 and 10 with plasmid pSG5. The PCR product was cut with Xmal and BamHI, purified by electrophoresis on a 2% agarose gel, isolated (e. g. with GENECLEAN, see above then ligated into the Xmal-BamHI cut pNG2 plasmid to give pNG3.

The pNG4 vector was prepared as follows. The pNG3 vector was further modified such that the Sad restriction enzyme recognition site in the cloned CMV promoter fragment was corrupted by changing the DNA sequence. This was achieved by the use of a two step PCR mutagenesis reaction using the pNG3 vector as a template. The PCR used two complementary oligonucleotide primers (SEQ ID NOS: 11 and 12) to mutate the Sac I recognition sequence and 2 flanking primers (SEQ ID NOS: 13 and 14) for product amplification. Two primer pairs (SEQ ID NOS: 11 and 14 and (SEQ ID NOS: 12 and 13) were used in a standard PCR reaction (as described above) to obtain the initial 2 PCR products, which were isolated by electrophoresis on 2% agarose gels. Equimolar amounts of each product were mixed and reamplified using the flanking primers (SEQ ID NOS: 13 and 14) under the standard PCR reaction conditions to splice together and amplify the final PCR Product. This product was subsequently digested with the restriction enzymes Ncol and Hindlll and cloned into the appropriately restricted and prepared pNG3 vector such that the mutated (Sad site minus) fragment replaced the original pNG3 Ncol-Hind III (Sad site plus) fragment. This new vector was named pNG4. A clone of the PQXB1/2 murine light chain in the Bluescript KS+ vector (VKPQ1) is taken and amplified using oligonucleotide primers specific for PQXB1/2 light chain. Similarly a PQXB1/2 heavy chain clone (VHPQ1) is amplified using oligonucleotide primers specific for PQXB1/2 heavy chain. The PCR is performed as follows: To lOOng of plasmid DNA are added 5 μl dNTPs (2.5mM), 5ul 1 Ox Enzyme buffer (see above), lul of 25pM/ul forward primer, lul of 25pM/μl back primer, 0.5 μl AmpliTaq (Perkin-Elmer cetus) and DEPC-treated water to obtain a volume of 50 μl. The PCR conditions are set for 15 cycles at 94 °C for 90s; 55 °C for 60s; 72 °C for 120s, ending the last cycle with a further 72 °C for 10 min incubation. The products are analysed on a 2% agarose gel. The DNA is purified (using GENECLEAN) and the DNA fragment digested with the relevant restriction enzymes in preparation for subsequent vector cloning.

For secretion of antibody light chain, a double stranded DNA cassette which contained both the information for a Kozak recognition sequence and a light chain signal sequence was designed. The cassette consisted of two individual oligonucleotides (SEQ ID NOS: 15 and 16) which were hybridised and subsequently cloned between the Hindlll and SacII restriction site of the pNG3 plasmid (which had been appropriately restricted and isolated using standard methodology) to create the vector pNG3-Vkss. The DNA sequence of SEQ ID NO: 19, which contains sequence for the human light chain kappa constant region, was digested with Xmal and Xhol and inserted between the Xhol and Xmal cut pNG- Vkss plasmid to give the vector pNG3-Vkss-HJuCk (NCIMB no. 40798). The light chain gene sequence described above is inserted, in frame, by cloning directly between the SacII and Xhol sites of the pNG3-Vkss HuCk vector. The PCR fragment obtained for the light chain gene is digested with SacII and Xhol restriction enzymes and cloned into the similarly restricted expression vector containing the VK signal and HuCK constant region coding sequences. The chimaeric PQXB1/2 light chain sequence created is named FABKPQ1. Similarly, for secretion of antibody heavy chain, a double stranded DNA cassette which contained both the information for a Kozak recognition sequence and a heavy chain signal sequence was designed. The cassette consisted of two individual oligonucleotides (SEQ ID NOS: 17 and 18) which were hybridised and subsequently cloned the Hindlll and EcoRI restriction site of the pNG4 plasmid (which had been appropriately restricted and isolated using standard methodology) to create the vector pNG4-VHss. The EcORl restriction enzyme recognition site may be required to be corrupted to allow the construction of the plant expression vector pPQkFabl . This is performed by changing the DNA sequence using a 2-step PCR mutagenesis protocol as described earlier. Heavy chain gene sequences can thus be inserted, in frame, by cloning directly between the EcoRI and Sad sites of the pNG4-VHss vector. The DNA sequence of SEQ ID NO: 20, which contains the coding sequence for human heavy chain IgG2CHl ' constant region (SEQ ID NOS: 23 and 24) was digested with Sa and Xmal and cloned into pNG4-VHss cut with Sad and Xmal to give the vector pNG4-VHss-HuIgG2CHl ' (NCIMB No. 40797). For Fab identification and purification puφoses sequences encoding the c-myc and 6-His motifs may be added, in frame, at the 3' end of the human heavy chain, IgG2CHl '. This is accomplished using overlapping oligonucleotide primers in a PCR mutagenesis protocol as described earlier. The PCR fragment obtained for the heavy chain gene was digested with EcoRI and Sad restriction enzymes and cloned into the similarly restricted expression vector pNG4-VHss- HuIgG2CHl ' containing the VH signal and HuIgG2 CHI ' constant region coding sequences. The chimaeric PQXB1/2 HuIgG2 Fd' chain sequence created is named FABHPQ1. In some instances it may be preferable to use other classes of chimaeric heavy chain

Fd' constructs. To this end, variants of the heavy chain vector are made containing HulgGlCHI ' (SEQ ID NOS: 21 and 22) or HuIgG3CHl ' (SEQ ID NOS: 25 and 26) which are substituted for the HuIgG2CHl ' (SEQ ID NOS: 23 and 24) gene.

Once the individual heavy and light chain sequences are constructed a heavy chain Fd' gene expression cassette (including both promoter and gene are excised as a Bglll/Sall fragment and cloned between into the BamHI/Sall sites of the light chain vector to produce a co-expression vector construct. This construct is transfected into NSO myeloma cells (ECACC No. 85110503) via standard techniques of electroporation and transfectants selected for the property of G418 resistance, a trait which is carried as a selectable marker on the expression plasmid construct.

Alternatively the complete heavy chain Fd' and light chain genes may simply be excised from their respective vectors as Hindlll/Xmal fragments and subsequently cloned into other expression vector systems of choice. Analysis of Mammalian Expression Products

Expression levels and molecular weight of PQXB1/2 Fab fragments are assessed by western blot analysis. Antibodies raised to constant regions of the Fab fragments or antibodies raised to the 6His or c-myc motifs are used to specifically identify heterologous Fab fragments. To determine whether the PQXB 1/2 Fab fragments bind conjugated or free paraquat, immunoassay techniques are employed as described below: -

1. Conventional paraquat antibody immobilised. Transformed eukaryotic cell line supernatants mixed with paraquat enzyme conjugate. Substrate added. Signal inversely proportional to fragment binding.

2. Fragments from transformed eukaryotic cell line supernatants captured using immobilised 6His or c-myc tag. Free paraquat or paraquat enzyme conjugate added. Substrate added. Signal proportional to fragment binding.

3. Free paraquat or paraquat protein conjugate immobilised. Supernatants added. Tag detection reagents added. Signal proportional to fragment binding.

Biomolecular interaction analysis (BIAcore) - Immunoassay formats 1-3 performed using surface plasmon resonance to monitor biomolecular interactions.

Affinity measurements of the heterologously expressed PQXB 1/2 Fab fragments for paraquat can be determined by immunoassay or BIAcore techniques as described below:- Immunoassay - Compare ability of fragments to bind paraquat to that of the conventional antibody e.g. IC₅₀ values. To make a comparison fragment concentrations are required. To enable this the 6His tag is utilised for purification puφoses using conventional techniques.

BIAcore - 1. Determine affinity values. Measurement of fragment concentration is also required. 2. Compare rates of dissociation (without purification) by immobilising equivalent amounts of antibody and fragments and determining off rates in the presence of paraquat.

Construction of Plant Expression Vector

The chimaeric PQXB 1/2 light chain, consisting of the Vk signal, PQXB 1/2 variable region of the light chain and the HuCK constant region coding sequences are PCR amplified from FABKPQ1 using flanking oligonucleotide primers that introduce sequences encoding the Ncol restriction site 5' to the signal sequence and a Smal restriction site 3' to the HuCK constant region. The Ncol/Smal digested PCR product is then ligated into the Ncol/Smal digested pUC based vector pMJBl (see Figure 1) to form pFabvkl. pMJBl is based on pIBT211 containing the CaMV 35S promoter with duplicated enhancer linked to the Tobacco Mosaic Virus translational enhancer 'omega' sequence replacing the tobacco etch virus 5' non-translated leader, and terminated with the nopaline synthase poly (A) signal (nos 3').

The chimaeric PQXB 1/2 heavy chain consisting of the VH signal, PQXB 1/2 variable region of the heavy chain and HuIgG2 CHI ' constant region coding sequences fragment is PCR amplified from FABHPQ1 in the same manner and ligated into pMJBl to form pFabvhl. Competent E. coli cells (DH5α) are transformed with these plasmids by a heat shock method. They are grown on L-agar and kanamycin plates. Positive colonies are checked by PCR or by hybridisation with labelled probes.

The fragment containing the enhanced CaMV 35S promoter, the TMV enhancer, the chimaeric PQXB 1/2 light chain and the NOS terminator is isolated from pFabvkl using Hind III and EcoRI . This fragment is ligated into EcoRl/Hind III cut JRli (Figure 2) to generate a Binl9 based plant transformation vector pPQkFabl.

The fragment containing the enhanced CaMV 35S promoter, the TMV enhancer, the chimaeric PQXB 1/2 heavy chain and the NOS terminator is isolated from pFabvhl using EcoRI and Hind III, the EcoRI site being converted to a Hind III sticky end by the ligation of oligonucleotide adapters, and the resultant fragment is ligated with Hindlll cut pPQkFabl to produce pPQhkFabl .

The orientation of the insert into pPQkFabl is checked by sequencing the DNA at the vector/insert junction.

Construction of a further Expression Vector

The Arabidopsis polyubiquitin promoter UBQ3 is isolated from pDmAMPULC (Figure 3) as a Hind III/Xhol fragment. pDmAMPULC is based on JRli where the CaMV 35S promoter is replaced with the Arabidopsis polyubiquitin promoter UBQ3. pFabvkl is treated with Hind III/Xhol to release the enhanced CaMV 35S promoter. The Hind III/Xhol promoter fragment of UBQ3 is then ligated into this cassette to produce pFabvk2. The fragment containing the UBQ3 promoter, the TMV enhancer, the chimaeric PQXB 1/2 light chain and the NOS terminator is isolated from pFabvk2 using Hind III and EcoRI . This fragment is ligated into EcoRl/Hind III cut JRli (Figure 2) to generate a Binl9 based plant transformation vector pPQkFab2. The chimearic PQXB 1/2 heavy chain is excised from pFabvh2 with Hindlll as described above and ligated with Hindlll digested pPQkFab2 to produce the plant expression vector pPQhkFab2. This vector contains the PQXB 1/2 light chain expressed from the UBQ3 promoter and the PQXB 1/2 heavy chain expressed from the CaMV 35S promoter.

Plant Transformation

Plasmids pPQhkFabl and/or pPQhkFab2 are transferred into Agrobacterium tumefaciens LBA4404 using the freeze/thaw method described by Holsters et al. (1978) Mol. Gen Genet 163. 181-187.Tobacco {Nicotiana tabacum cv. Samsun) transformants are produced by the leaf disc method as described by Bevan (1984), Nucleic Acids Res. L2 8711-8721). Shoots are regenerated on medium containing 1 OOmg/1 kanamycin. After rooting, plantlets are transferred to the glasshouse and grown under 16 h light/ 8 h dark conditions.

Polymerase Chain Reaction Analysis

Genomic DNA for polymerase chain reaction (PCR) analysis of transgenic plants is prepared according to Edwards et al. (1992) Nucleic Acids Res. 19f6 1349. PCR is performed using the methods described by Jepson et al. (1991) Plant Mol. Biol. Reporter 9(1) 131-138. The presence of the genes is determined using primers for internal regions of the genes encoding the Fab fragments and primers encoding sequence from the CaMV35 S and UBQ3 promoters.

Western Blot Analysis To verify the heterologous expression of the Fab fragments in tobacco western blot analysis is performed. A soluble protein extract is prepared from tobacco leaf and samples are loaded onto an SDS-PAGE gel and the proteins separated electrophoretically under denaturing conditions. The proteins are then electroblotted onto nitrocellulose and incubated with antibodies specific for the 6His or c-myc motif or with antibodies raised against the constant regions of the Fab fragment. Immunodetection is completed with the use of an anti-rabbit antiserum as second antibody, associated with the horseradish peroxidase and by enhanced chemical luminescence detection. Southern Blot Analysis

The pattern of integration of transgenes is verified by Southern Blot analysis. Genomic DNA is isolated from fresh tobacco leaves taken from plants in the glasshouse and digested with suitable restriction enzymes. Fragments are electrophoretically separated in 0.8% agarose and transferred onto nylon membrane by capillary blotting. The DNA is then hybridized with probes labelled with ³²P dCTP using the random priming protocol described by Feinberg and Vogelstein (1984) Anal. Biochem.J37 266-267 and (1983) Anal. Biochem. 132 6-13.

Functionality of FAB Fragments expressed in Transgenic Tobacco. The functionality of Fab fragments expressed in the cell walls of plant tissue is determined by measuring the ratio of free and bound paraquat and by studying the proportion of total paraquat sequestered in the cell wall fraction.

Paraquat sprayed plants or spiked plant extracts are separated into their cellular fractions by ultracentrifugation and the ratio of free to bound paraquat determined by radiolabel, spectrophotometric, capillary electrophoresis or immunoassay techniques.

Paraquat Spray Trials

30-50 primary transformants, growing on MS medium, are produced and analysed by PCR for the presence of the transgenes. Two explants, consisting of a leaf attached to a piece of shoot are transferred onto fresh MS media. One of each pair is transferred into soil once a root system has developed and grown in the glasshouse for seed production. The other is transferred into MS medium containing the lowest concentration of paraquat that results in phytotoxic symptoms in wild-type tobacco. Transformants that do not display injury symptoms are grown further to provide enough explants for a dose response curve to be determined for paraquat injury. Selected lines are then grown to provide enough explants for glasshouse spray trials. 20 plants for each transformed line are transferred into soil and sprayed with a range of paraquat concentrations using conventional techniques and the degree of tolerance assessed.

Production of Homozygous Lines

Explants that are transferred to soil and grown to flowering in the glasshouse are self pollinated. Samples of the seed produced are used for segregation analysis by germinating seed on 1/2 MS-media containing lOOmg/1 kanamycin. Seeds are grown for 1-2 weeks prior to scoring.

(A) Production of the fused variable heavy (Vh) and variable light (V\) domains using fusion PCR

The DNAs encoding the variable light and heavy domains (depicted as SEQ ID Nos. 1 and 3 respectively) are fused using complementary Polymerase Chain Reaction (PCR) primers, the complementary regions encoding the flexible protein linker sequence depicted as SEQ ID No. 27. The PCR technique involved two classical PCR steps each amplifying the variable heavy depicted as SEQ ID Nos. 31 and 32, or Variable light depicted as SEQ ID Nos. 33 and 34, sequences using primers possessing "overhangs" which allow cloning and/or encoding the protein linker.

The two PCR products were mixed, heat denatured and used as the "template DNA" in a third PCR reaction using Vh5'NcoI (SEQ ID No. 31) and VB'Xmal (SEQ ID No. 34) as the PCR primers.

(B) Production of an Expression vectors containing the linked Vh and VI domains.

The resulting PCR products obtained via the method of part (A) were digested with the restriction enzymes Ncol and Xmal and then cloned into one of three vecors denoted as pPRS, pDm (Figure 4) and pPG. These vectors each comprise a double 35S CaMV enhancer and promoter, a sequence encoding a signal peptide (where pPRS contains the Tobacco PR-S signal sequence depicted as SEQ ID No. 30, pDm contains the Dahlia AFP signal sequence depicted as SEQ ID No. 28 and pPG contains the Tomato PG signal sequence depicted as SEQ ID No. 29), a hex histidine tag and a functional terminator sequence.

(O Production of plant transformation vectors.

The expression cassettes were excised from the expression vectors according to (B) above by restriction digestion with Hindlll and EcoRI and then ligated into the plant transformation vector pBinl9 (Figure 5). Each of the three plant transformation vectors (differing only in the type of signal sequence that they contained) were used to transfect Agrobacterium tumefaciens strain LB4404 using standard electroporation techniques. These transfected Agrobacterium were then used to transform Tobacco (variety Samsun) using a standard leaf disk technique-

(D) Analysis of plants transformed with the pBin 19 vector containing the pPG expression cassette.

Plants transformed according to (C) above were analysed for gene insertion by PCR and for protein level by Western blotting. Plants were then clonally propogated in tissue culture using internode sections to get 10 transformed plants of each line. The lines were transferred to soil and grown to 10cm in height. The plants were then sprayed with Paraquat solutions of concentration range 0-400g/Ηa. Control plants were represented by untransformed tobacco from tissue culture that were transferred to soil at the same time as the transformants. These "wild type" plants were also sprayed with Paraquat solutions of concentration range 0- 400g/Ha.

The line expressing the highest level of antibody viz. 4% total soluble protein, survived the highest Paraquat solution treatment.

Bacterial Expression and Analysis of the fused Vh and VI domains

The PCR fusion products obtained via the methods described in Example 3 (A) were also cloned into a bacterial expression vector and the protein was produced and harvested using standard techniques. The binding capacity of the protein was compared to that of a mutant sequence published by Porter et al (1995) J. Chem. Technol. Biotechnol. July 63(3) pp279- 289 titled "Cloning Expression and Characterisation of a single-chain antibody fragment to the herbicide Paraquat". The protein according to the present invention bound to Paraquat with 100 fold greater efficiency. The results of the comparison analysis are shown in Figure X. PQ-ZEN1 scAb denotes the protein produced according to the methods of Example 4 and BG1 scAb denotes the protein sequence which was partially published in the Porter et al (1995) reference.

Claims

1. A protein which is capable of binding a herbicide, said protein comprising the sequence depicted as SEQ ID No 2 or an active part thereof or SEQ ID No 4 or an active part thereof with the proviso that the active part is not that depicted as SEQ ID Nos. 35 to 38.

2. A protein according to claim 1 characterised in that the herbicide is paraquat or diquat or a structurally related analogue thereof.

3. A protein which is capable of binding a herbicide said protein comprising the amino acid sequence depicted as SEQ ID No 2 or an active part thereof or SEQ ID No 4 or an active part thereof wherein the said active part has a K_a greater than 1.24 x 10^╬┤ liters/mole for paraquat.

4. A protein which is capable of binding a herbicide, said protein comprising the sequence depicted as SEQ ID No 4 or an active variant thereof comprising the region shown in SEQ ID No 4 at amino acid number 35 to 45 or amino acid number 25 to 35 or amino acid number 26 to 36 or amino acid number 27 to 37 or amino acid number

28 to 38 or amino acid number 29 to 39 or amino acid number 30 to 40 or amino acid number 31 to 41 or amino acid number 32 to 42 or amino acid number 33 to 43 or amino acid number 34 to 44 or any conservative substitution of any amino acid within said active variant with the proviso that the resulting K- of the protein or variant for paraquat is greater than 1.24 x 10⁶ liters/mole.

5. A protein which is capable of binding a herbicide, said protein comprising the sequence depicted as SEQ ID No 2 or an active part thereof which is linked to the sequence depicted as SEQ ID No 4 or an active part thereof, said sequences optionally comprising cysteine residues disposed in such a way as to facilitate the formation of disulphide cross bridges between them.

6. A protein according to claim 5 characterised in that the said sequences are linked via a synthetic linker.

7. A protein according to claim 6 characterised in that the said synthetic linker has the sequence depicted as SEQ ID No. 27.

8. A protein according to any one of claims 5 to 7 characterised in that the herbicide is paraquat or diquat or a structurally related analogue thereof.

9. A polynucleotide sequence encoding the protein or active part or variant according to any one of claims 1 to 8.

10. A polynucleotide according to claim 9 comprising the sequence depicted as SEQ ID No l.

11. A polynucleotide according to claim 9 comprising the sequence depicted as SEQ ID No 3.

12. A polynucleotide sequence which is the complement of one which binds to a polynucleotide according to any one of claims 9 to 11 at a temperature of between

60┬░C and 65┬░C in 0.3 strength citrate buffered saline containing 0.1% SDS followed by rinsing at the same temperature with 0.3 strength citrate buffered saline containing 0.1% SDS and which still encodes a protein which is capable of binding a herbicide, with the proviso that the polynucleotide sequence does not encode a protein which has a K. of less than or equal to 1.24 x 10⁶ liters/mole for paraquat.

13. A polynucleotide sequence according to claim 12 wherein the said herbicide is paraquat or diquat or a structurally related analogue thereof.

14. A DNA construct comprising in sequence a promoter region which is operable in plants operably joined to a polynucleotide sequence according to any one of claims 9 to 13 and a transcription termination region.

15. A DNA construct according to claim 14 further comprising (a) transcriptional enhancing elements; and/or (b) regions encoding non translated translational enhancing sequences such as Omega and Omega prime; and/or (c) regions encoding non translated sequences such as intron sequences; and/or (d) regions encoding target sequences which are capable of directing transcription products to either intracellular organelles, intracellular compartments, cell membranes or to the outside of the cell.

16. A DNA construct according to claim 15 characterised in that the regions encoding the said target sequences are selected from the group depicted as SEQ ID Nos. 40 to 47.

17. A DNA construct according to claim 15 characterised in that the said regions encoding the target sequences encode any one of the sequences depicted as SEQ ID Nos. 28 to 30.

18. A DNA construct according to any one of claims 14 to 17 wherein the promoter is constitutive.

19. A DNA construct according to any one of claims 14 to 17 wherein the promoter region is developmentally regulated.

20. A DNA construct according to claim 18 or claim 19 wherein the promoter is tissue specific.

21. A DNA construct according to claim 20 wherein the promoter is maize flag leaf specific.

22. A DNA construct according to any one of claims 14 to 21 which further comprises a polynucleotide sequence which is capable of imparting to a plant, substantial resistance or substantial tolerance to a further herbicide.

23. A DNA construct according to claim 22 wherein the said herbicide is N- phosphonomethylglycine or an agriculturally acceptable salt or ester thereof.

24. A plant transformation vector comprising a DNA construct according to any one of claims 14 to 23.

25. A method of producing plants which are capable of producing herbicide binding proteins comprising inserting into a plant cell a polynucleotide according to any one of claims 9 to 13 or a construct according to any one of claims 14 to 23 or a plant transformation vector according to claim 24, regenerating mo╧åhologically normal fertile plants or plant parts therefrom and selecting from the population of regenerants those mo╧åhologically normal fertile plants or plant parts which are capable of producing the said herbicide binding proteins.

26. Mo╧åhologically normal fertile plants and plant parts produced according to the method of claim 25.

27. Mo╧åhologically normal fertile plants according to claim 26 characterised in that said mo╧åhologically normal fertile plants are small grain cereals, oil seed crops, fibre plants, fruit, vegetables, plantation crops and trees.

28. Mo╧åhologically normal fertile plants according to claim 27 characterised in that plants are selected from the group consisting of soybean, cotton, tobacco, sugarbeet, oilseed rape, canola, flax, sunflower, potato, tomato, alfalfa, lettuce, maize, wheat, sorghum, rye, bananas, barley, oat, turf grass, forage grass, sugar cane, pea, field bean, rice, pine, poplar, apple, grape, citrus or nut plants and the progeny and seeds thereof.

29. A method of selectively controlling weeds in a field said field comprising weeds and crop plants said method comprising contacting said field paraquat, characterised in that the said crop plants are the plants according to any one of claims 26 to 28.

30. A method according to claim 29, further comprising application to the field of a pesticidally effective amount of one or more of an insecticide, fungicide, bactericide, nematicide and anti-viral.

31. Use of a polynucleotide according to any one of claims 9 to 13 or a construct according to any one of claims 14 to 23 or a plant transformation vector according to claim 24, in the production of mo╧åhologically normal fertile plants which are substantially resistant or tolerant to herbicides which inhibit photosynthesis by accepting electrons from photosystem I thus generating free radicals which cause lipid peroxidation, or (ii) by blocking electron transport in photosystem II.

32. Use according to claim 31, wherein the herbicide is paraquat or diquat or a structurally related analogue thereof.

33. Use of a polynucleotide comprising a protein encoding region having the sequence depicted as SEQ ID No. 1 and/or SEQ ID No. 3 as a selectable or screenable marker gene in the selection or screening of transgenic plant material.

34. An open reading frame coding for a functional paraquat binding protein having at least 99% probability of being the same as SEQ ID No. 1 or SEQ ID No. 3 as recognised by the FASTA algorithm.

35. An open reading frame coding for a functional paraquat binding protein having a smallest sum probability of 5 x 10^"45 of being the same as SEQ ID No. 1 or SEQ ID No. 3 as recognised by the BLASTP 2.0al OMP-WashU algorithm.

36. An open reading frame coding for a variable heavy antibody domain which can be used to produce a functional paraquat binding protein being 76-100% identical to SEQ ID No. 4 according to the Needleman and Wunsh equation.