Monsanto Technology LLC v Cargill International SA & Anor

This is an action for infringement of EP (UK) 0 546 090 which stands in the name of the claimants (‘Monsanto’). The invention is entitled ‘glyphosate tolerant 5-enolpyruvylshikimate-3-phosphate synthases’ and is concerned with enzymes which, if expressed in a plant, confer resistance to the herbicide Round Up, which otherwise inhibits the so-called shikimate pathway in the synthesis of certain of the aromatic amino acids. The enzymes are referred to in the specification as EPSPSs.

The defendants (‘Cargill’) buy soya beans grown from seed carrying the gene for one of the EPSPSs disclosed in the patent in Argentina, and from it they or others manufacture meal which they import into the United Kingdom. The cargo in issue in this action is 5000 tonnes shipped on the MV Podhale. The evidence was that the use of seed carrying a gene for an enzyme called CP4R, generally referred to as ‘Round Up Ready’ (‘RuR’) seed, had transformed the soya bean industry in Argentina, and represented a very substantial benefit to Argentinean growers and processors. I was told that 99% of soya bean meal exported from Argentina came from Round Up Ready plants carrying the CP4R gene. Given the structure of the claims, Cargill say that there is no infringement of the UK part of this patent. Validity is challenged on the grounds of anticipation, obviousness and insufficiency although Cargill accept that the discovery of the CP4 enzyme was an invention.

Anticipation is alleged on the basis of a number of disclosures of EPSP synthases. Most important is that of Bacillus subtilis (‘B subtilis’ or just ‘subtilis’) said to be disclosed in a paper called Henner; and that of strain PG2982 of Pseudomonas aeruginosa (‘P aeruginosa’) disclosed in a paper called Moore. Obviousness is alleged over Henner and Moore, and over a document called Fischer. A point of some general importance arises in relation to B subtilis EPSPS, because Cargill did no experiments of its own to show that that enzyme reacted with antibodies raised against the enzymes specified in claim 1. Without complying with the order for notices of experiments, it sought to rely on certain experiments performed on behalf of a third party, Syngenta, for use in opposition proceedings in the EPO. Kitchin J adjourned an application to admit these experiments, and I have to decide whether to admit them. A point of some interest also arises on the use of the contents of the EPO file in construing the specification.

The priority date of the patent is 1990. The analytical techniques used in biochemistry and the techniques of genetic engineering which would be employed by the team of skilled scientists seeking to put this patent in effect have been described in detail in a number of judgments and I do not propose to set them out at length. So far as analytical techniques are concerned, for the genetic material there was a great forensic dispute as to the proper interpretation of Southern blots, and much time was spent on the interpretation of Western blots so far as the enzymes were concerned. No particular reference will be made to the general techniques of genetic engineering. One minor issue arises concerning the sequence encoding a chloroplast transit peptide which ensures that nucleus-expressed EPSPS finds its way to the chloroplast, where it is active. On the technical issues, Professor Bartlett, an expert in enzyme kinetics and Professor Lichtenstein, a molecular biologist, gave evidence on behalf of Monsanto. Professor Leaver, a plant biologist, and Dr Isaac, a molecular biologist, gave evidence for Cargill. Dr Isaac is the only witness to be subject to sustained criticism (Professor Lichtenstein is said to have been too much of an advocate) and it is convenient to deal with those criticisms where they arise.

Monsanto apply to amend the patent in suit. The allowability of the amendments is disputed, and if they are allowable it is said that the court should not exercise its discretion to permit them to be made. On this issue Dr Padgette, one of the inventors and Dr von Menges, who prosecuted the patent through the EPO, gave evidence. They are not criticised as witnesses.

The only other substantive issue raised by Cargill was an allegation that Monsanto was estopped by its acquiescence in the acts of importation complained of. This issue necessitated the examination of some of the commercial developments in Argentina, where Monsanto has no protection for Round Up Ready soya beans. On this issue, Mr Tobin gave evidence for Monsanto, and Mr Amado for Cargill. Mr Amado was an excellent witness.

This specification is concerned with an improvement to crops by way of making them resistant to glyphosate herbicide. To make a crop resistant to glyphosate permits destruction of weeds by use of glyphosate among the growing plants, so reducing the need for tillage. There was a measure of agreement that it is likely that the patent would be addressed to a team, including a post-doctoral molecular biologist. By 1990, I think that it was established that the techniques employed in the specification were normal in this area of research, and I shall return to this issue as necessary.

At the priority date, the mode of operation of glyphosate, better known under the brand ‘Round up’, was well known. Glyphosate (which is N-phosphonomethylglycine) is an extremely successful herbicide, much of its success being attributable to its lack of effect on animals, its rapid decomposition to harmless products in the earth and its drastic effect on plants. Its mode of operation is by curbing the conversion of phosphoenolpyruvic acid (‘PEP’) and 3-phosphoshikimic acid to 5-enolpyruvyl-3-phosphoshikimic acid by inhibiting the enzyme 5-enolpyruvylshikimate synthase (‘EPSP synthase’). This enzyme is essential to aromatic amino acid biosynthesis in microorganisms and plants. The aromatic amino acids are essential components of proteins. It seems that EPSPSs differ from organism to organism to a greater or lesser extent.

Specific enzymes catalyse specific chemical reactions, and possess an ‘active site’ at which one of the reagents will bond during the reaction. This reagent is called the enzyme’s substrate. Competitive enzyme inhibitors work by binding to the enzyme’s active site in preference to the substrate, so blocking the reaction. There are two ways of getting over this: make much more enzyme, or find a different one which will still catalyze the reaction of interest but will not bind to the inhibitor.

At the priority date, the necessary techniques for identifying the enzyme of interest, for characterising it (including immunological techniques), for finding the DNA that encoded it, for transforming plants and for expressing the recombinant protein in their progeny were well established. The techniques are anything but straightforward, and they take, potentially, a lot of time. This potentially places a difficult burden on the draftsman of the specification who must identify with some precision which techniques and what conditions he says should be used. It also means that allegations of insufficiency have to be treated with caution.

The specification starts with a summary of the mode of operation of glyphosate.

The specification acknowledges that transgenic plants have been produced ([0005]) in which a higher level of EPSP synthase has been produced, and in which the EPSP synthase has itself been modified. The key characterisation of these EPSPSs is as follows, references omitted:

‘It has been shown that glyphosate tolerant plants can be produced by inserting into the genome of the plant the capacity to produce a higher level of EPSP synthase in the chloroplast of the cell…which enzyme is preferably glyphosate tolerant. Variants of the wild-type EPSPS enzyme have been isolated which are glyphosate tolerant as a result of alterations in the EPSPS amino acid coding sequence. These variants typically have a higher K_i for glyphosate than the wild-type EPSP enzyme which confers the glyphosate tolerant phenotype (Footnote: ¹), but these variants are also characterised by a high K_m for PEP which makes the enzyme kinetically less efficient.’

This passage necessitates a short diversion into reaction kinetics.

Absent glyphosate, the crucial step in the shikimate pathway is for present purposes the binding of PEP, the substrate, to EPSPS. A reaction catalysed by an enzyme can be characterised by the maximum rate at which a single molecule of the enzyme can process molecules of substrate. General considerations show that

Glyphosate competes with PEP to form a complex with the enzyme, the enzyme/inhibitor complex, EI. It can be shown that the apparent K_m value is changed to , where

K_i (glyphosate) is again a concentration. It is the concentration of glyphosate at which one-half the available sites on the EPSPS are occupied.

It follows that the ideal glyphosate-resistant EPSPS has a low K_m (PEP) to maximise the rate of conversion of PEP, and a high K_i (glyphosate) showing that it does not bind effectively with glyphosate. Paragraph [0005] gives some examples: native E. coli: K_m(PEP) 10µM, K_i (glyphosate) 0.5µM; the same, with a point mutation at position 96: K_m(PEP) 220µM and K_i (glyphosate) 4.0mM. The latter is highly glyphosate-tolerant, but K_m(PEP) is too high, and so the EPSPS is not efficient. Paragraphs [0006] and [0007] say that no such EPSPS has been discovered thus far.

Paragraphs [0008]-[0010] are of great importance. I will return to them, but set them out here with some commentary.

‘[0008] A DNA molecule comprising DNA encoding a kinetically efficient, glyphosate tolerant EPSP synthase is presented. The EPSP synthases of the present invention reduce the amount of over production of the EPSPS enzyme in a transgenic plant necessary for the enzyme to maintain catalytic activity while still conferring glyphosate tolerance.’

This statement indicates that K_m is low, and this means that a transgenic glyphosate tolerant plant need not incorporate the wherewithal to express much EPSPS in the chloroplast.

‘This and other EPSP synthases described herein represent a new class of EPSPS enzymes, referred to hereinafter as Class II enzymes. Class II EPSPS enzymes share little homology to known bacterial or plant EPSPS enzymes and exhibit tolerance to glyphosate while maintaining suitable K_m(PEP) ranges. Suitable ranges of K_m(PEP) for EPSPS for enzymes of the present invention are between 1-150µM, with a more preferred range of between 1-35µM and a most preferred range between 2-25µM. These kinetic constants are determined under the assay conditions specified hereinafter.’

These figures are really only a statement of the general desire for a comparatively low K_m(PEP). No method of determining K_m is in fact given in the specification, and, since two methods (phosphate release and radioactive HPLC both mentioned in a cross-reference to a paper of Dr Padgette’s at paragraph [0042]) will give different answers the specification is said to be insufficient.

‘The V_max of the enzyme should preferably be at least 15% of the uninhibited plant enzyme and more preferably be greater than 25%. An EPSPS of the present invention preferably has a K_i for glyphosate range between 25-10000 µM. The K_i/K_m ratio should be between 3-500, and more preferably between 6-250. The V_max should preferably be in the range of 2-100 units/mg (µmoles/minute.mg at 25ºC) and the K_m for shikimate-3-phosphate should preferably be in the range of 0.1 to 50µM.’

As I understand it, the numerical ranges in this passage do not introduce any effective limitation on the description of the invention.

The patent then gets down to business, if I may put it that way, by describing the discovery of three genes coding for EPSPSs with the desirable characteristics described.

‘[0009] Genes coding for Class II EPSPS enzymes have been isolated from three (3) different bacteria: Agrobacterium tumefaciens sp strain CP4, Achromobacter sp. strain LBAA, and Pseudomonas sp. strain PG2982. The LBAA and PG2982 Class II EPSPS genes have been determined to be identical and the proteins encoded by these two genes are very similar to the CP4 protein and share approximately 84% amino acid identity with it. ’

The specification makes it clear subsequently that it is only reporting work done on these three genes. Although these are not the only genes identified as expressing Class II EPSPSs, they encode the only enzymes characterised in the specification. Accordingly the characterising features of the Class II EPSPS are important:

‘Class II EPSPS enzymes can be readily distinguished from Class I EPSPS’s by their inability to react with polyclonal antibodies prepared from Class I EPSPS enzymes under conditions where other Class I EPSPS enzymes would readily react with the Class I antibodies.’

Class I antibodies are not explicitly described, but the inference is that the class of enzymes will have a large number of members: essentially all EPSPSs from glyphosate-sensitive plants. The specification adds some substance to the suggestion that the Class I and Class II enzymes are in some way different in principle by comparing the proteins (see in particular [0034]) and encoding sequences—see [0069]—but that it as far as it gets. There is no description of any raising of antibodies (specifying what animal or animals to be used, how many animals, which class I material is to be used as antigen, how the antigen is to be administered and so on). The one description of an immunological test is in paragraph [0034] in which Petunia EPSPS rabbit antibody is reacted with crude bacterial extract CP4 in the presence of petunia EPSPS as a control, for the purpose of showing different reactivities with the rabbit antibody. The test prescribed by the claim is different, and I shall return to this topic below.

‘[0010] Other Class II EPSPS enzymes can be readily isolated and identified by utilizing a nucleic acid probe from one of the Class II EPSPS genes disclosed herein using standard hybridization techniques. Such a probe from the CP4 strain has been prepared and utilized to isolate the Class II EPSPS genes from strains LBAA and PG2982. These genes may also be adapted for enhanced expression in plants by known methodology. Such a probe has also been used to identify homologous genes in bacteria isolated de novo from soil.’

The hybridization conditions are described in [0060]-[0065].

‘[0011] The Class II EPSPS enzymes are preferably fused to a chloroplast transit peptide (CTP) to target the protein to the chloroplasts of the plant into which it may be introduced. Chimeric genes encoding this CTP-Class II EPSPS fusion protein may be prepared with an appropriate promoter and 3’ polyadenylation site for introduction into a desired plant by standard methods.’

Paragraph [0012] is explicit in its statement that the invention lies in the new class of enzymes:

‘[0012] Therefore, in one aspect, the present invention provides a new class of EPSP synthases that exhibit a low K_m for phosphoenolpyruvate (PEP), a high V_max/K_m ratio, and a high K_i for glyphosate such that when introduced into a plant, the plant is made glyphosate tolerant such that the catalytic activity of the enzyme and plant metabolism are maintained in a substantially normal state. For the purposes of this discussion, a highly efficient EPSPS refers to its efficiency in the presence of glyphosate.’

After a brief discussion of the figures, the specification moves onto the identification of highly tolerant EPSPS enzymes. So far as sequences coding for enzymes of interest in the invention are concerned, the figures show structural DNA sequences for CP4, LBAA, and PG2982, and a synthetic gene for CP4 (Sequence ID No 9; Figure 8). The cloning of CP4 is described in detail, as is its fusion to a gene encoding a chloroplast transfer protein.

A passage entitled ‘Alternative Isolation Protocols for Other Class II EPSPS Structural Genes’ starts at paragraph [0066]. Further examples of Class II enzymes or of genes encoding them are given briefly in [0068], but little detail is provided. What is said is that the difficulty in obtaining the Class II gene was caused by the lack of similarity between Class I and Class II enzymes, and that the CP4 gene probe greatly facilitated the identification of the other genes (LBAA and PG2982, which are in fact the same). Accordingly, ‘[t]he high degree of similarity of the Class II EPSPS genes may be used to identify and clone additional genes in much the same way that Class I EPSPS gene probes have been used to clone other Class I genes’.

The passage entitled ‘Relationships between EPSPS genes’ (paragraphs [0069] ff.) deals only with the difference between CP4 EPSPS from a structural point of view against other EPSPSs from bacteria, yeast and plants (Class I) and with LBAA (Class II). After this general description, substantially the whole of the rest of the document is concerned briefly with the construction of a synthetic CP4 gene thought to have advantages in expression, and the actual transformation of plants to include this gene in their genome. In [0077], which is otherwise introductory in nature, it is stated that it was a by-product of the necessary fusion of the gene for chloroplast transit peptide (CTP) that the second codon of the CP4 gene was converted to one for leucine. This eventually became the Round Up Ready gene, and is the gene present in every soya plant in Argentina, among other places.

It is a curiosity of this specification that although the Round Up Ready gene does not encode CP4 as found in Agrobacterium sp. but a protein differing from it by a single mutation, serine to leucine at position 2, there is no explicit claim either to this EPSPS or to a sequence encoding it. The mutation is a side effect of the manner in which the chimeric gene used to transfect the parent plant was constructed.

With that summary of the disclosure, I can turn to the claims. They fall into groups. The claims alleged to be infringed are claims 1 to 5 (DNA sequences encoding for an enzyme); claims 7, 8, 11 and 12, which are concerned with the recombinant double-stranded DNA material; 14, 17 and 18 (method claims for producing genetically transformed plants).

At this point also, I can introduce the amended claims. Of these claims, 1, 2, 4 and 5 are alleged to be infringed. I am reluctant to set them out twice and have adopted the conventional approach of showing deletions by striking-through and additions by italics. Some claims are deleted in their entirety.

Although the amendments are sought unconditionally, Monsanto say that the existing claims are both valid and infringed. Accordingly I must decide whether these are amendments which may lawfully be made, and decide whether to permit the amendments in the exercise of the Court’s discretion.

Existing claims 1–5, new claims 1 and 2

‘1. An isolated DNA sequence encoding a Class II EPSPS enzyme,

said enzyme being an EPSPS enzyme having a K_m for phosphoenylpyruvate (PEP) between 1–150µM and a K_i(glyphosate)/K_m(PEP) ratio between 3–500,

which enzyme is capable of reacting with antibodies raised against a class II EPSPS enzyme selected from the group consisting of the enzymes of SEOSEQ ID NO: 3 and SEQ ID NO 5.

which DNA sequence encodes the amino acid sequence of SEQ ID NO:3 save that serine at position 2 is replaced by leucine.

2.

A DNA sequence of claim 1 wherein said K_m for phosphoenolpyruvate is between 2–25µM.

3.

A DNA sequence of claim 1 wherein said K_i/K_m ratio is between 6–250.

42. An isolated DNA sequence encoding a protein which exhibits EPSPS activity wherein

said protein is capable of reacting with antibodies raised against a Class II EPSPS enzyme selected from the group consisting of the enzymes of SEQ ID NO:3 and SEQ ID NO:5,

which DNA sequence encodes the amino acid sequence of SEQ ID NO:3 save that serine at position 2 is replaced by leucine.

5.

The DNA sequence of claim 4 wherein said antibodies are raised against a Class II EPSPS enzyme of SEQ ID NO:3.

63.

An isolated DNA sequence encoding a Class II EPSPS enzyme selected from the group consisting of SEQ ID NO: 3 and SEQ ID NO:5.

74.

A recombinant double-stranded DNA molecule compnsing comprising in sequence

a)

a promoter which functions in plant cells to cause the production of an RNA sequence selected from the group consisting of CaMV35S and FMV35S promoters;

b)

a structural DNA sequence that causes the production of an RNA sequence which encodes a fusion polypeptide comprising an amino-terminal chloroplast transit peptide and a class II EPSPS enzyme capable of reacting with antibodies raised against a Class II EPSPS enzyme selected from the group consisting of the enzymes SEQ ID NO:3 and SEQ ID NO:5 and

c)

a 3 non-translated region which functions in plant cells to cause the addition of a stretch of polyadenylnucleotides to the 3 end of the RNA sequence, being the NOS 3 polyadenylation sequence fragment

where the promoter is heterologous with respect to the structural DNA sequence and adapted to cause sufficient expression of the fusion polypeptide to enhance the glyphosate tolerance of a plant cell transformed with said DNA molecule and where the fusion polypeptide has the amino sequence of SEQ ID NO:15 fused through the common terminal methionine to the amino acid sequence of SEQ ID NO:3 in which serine at position 2 is replaced by leucine.’

None of the features of claims 8–13, which are ultimately dependent on claim 7, are incorporated by amendment into old claim 7 (new claim 4) with the exception of claim 8 (the chloroplast transit peptide), claim 12 (the two promoters) and claim 13 (encoding for EPSPS of SEQ ID NO:3). The reference in the amendments to SEQ ID NO:15 is to the Petunia-derived chloroplast transit peptide CTP4. Since these claims are all sought to be deleted, I shall not set them out here.

Claim 14 (new claim 5) is a method claim which matches the recombinant double-strand claim 7 (new 4). It is as follows:

‘145 A method of producing genetically transformed plants which are tolerant toward glyphosate herbicide, comprising the steps of:

a)

inserting into the genome of a plant cell a recombinant, double-stranded DNA molecule comprising:

i)

a promoter which functions in plant cells to cause the production of an RNA sequence, selected from the group consisting of CaMV35S and FMV35S promoters,

ii)

a structural DNA sequence that causes the production of an RNA sequence which encodes a 5 fusion polypeptide comprising an amino terminal chloroplast transit peptide and a Class II EPSPS enzyme capable of reacting with antibodies raised against a Class II EPSPS enzyme selected from the group consisting of the enzymes of SEQ ID NO:3 and SEQ ID NO:5,

iii)

a 3 non-translated DNA sequence which functions in plant cells to cause the addition of a stretch of polyadenyl nucleotides to the 3end of the RNA sequence being the NOS 3 polyadenylation sequence fragment

where the promoter is heterologous with respect to the structural DNA sequence and adapted to cause sufficient expression of the fusion polypeptide to enhance the glyphosate tolerance of a plant cell transformed with said gene

and where the fusion polypeptide has the amino acid sequence of SEQ ID NO: 15 fused through the common terminal methionine to the amino acid sequence of SEQ ID NO:3 in which serine at position 2 is replaced by leucine;

b)

obtaining a transformed plant cell; and

c)

regenerating from the transformed plant cell a genetically transformed plant which has increased tolerance to glyphosate herbicide.’

The same observations mutatis mutandis as I have made in relation to claim 7 may also be made in respect of this claim. Again, I will not set out claims 15-19, whose deletion is sought.

Claim 20 (new claim 6) is a claim to the plant cell:

‘A glyphosate tolerant plant cell comprising a DNA molecule of Claims 8, 9, 12 or 13 4.’

The remaining claims do not need to be quoted. Claim 23 is a claim to particular species of plant cell (including soybean), and claim 24 a claim to the plant. Finally, claim 28 (new claim 10) covers a method for selectively controlling weeds in a field containing a crop plant, in which the stages include using a crop which is glyphosate tolerant as a result of transforming a plant with a double-stranded DNA molecule defined in the familiar manner, and spraying the field with glyphosate. There is an express claim to performing this method when the field is full of soybean plants.

Quite apart from the specifically technical issues that arise on these claims, there are a number of more general questions which may conveniently be considered at this point. It will be recalled that the acts of alleged infringement relate to the cargo of soybean meal on board MV Podhale, and that to be justiciable here for present purposes the infringement must occur within the jurisdiction. There is no real doubt that the meal, or a very substantial part of it, is produced from Round Up Ready soybeans in Argentina. The first question is whether importation of such meal is capable of infringing any of the method claims. The second question is raised by an allegation by Monsanto that the meal contains at least genomic fragments containing the whole of the Round Up Ready gene. If this is so, are any of the claims relating to genetic material infringed?

Section 60(1)(c) of the Act provides

"(1)

Subject to the provisions of this section, a person infringes a patent for an invention if, but only if, while the patent is in force, he does any of the following things in the United Kingdom in relation to the invention without the consent of the proprietor of the patent, that is to say -

...

(c)

where the invention is a process, he disposes of, offers to dispose of, uses or imports any product obtained directly by means of that process or keeps any such product whether for disposal or otherwise."

In Pioneer v Warner [1997] RPC 759 the Court of Appeal accepted the following submission:

‘ Mr Prescott QC and Mr Howe QC, for the defendants, submitted that the first step, as always, is to construe the claims in each patent. Here their construction is not in doubt. In 649 the process ends with the production of the father; in each of the others with the production of the master. They and they alone are the products obtained directly by means of the processes. Mr Prescott and Mr Howe accepted that a product can be further processed without losing its identity, in which event it remains the product obtained directly by means of the patented process. The question whether it has lost its identity depends on whether it no longer retains its essential characteristics. It is only for that purpose that a consideration of essential characteristics is appropriate. There is no free-standing "essential characteristics" test. Mr Prescott and Mr Howe further submitted that here there is no identity between the masters and the finished discs. Father, mother and son are each separate products with identities different from that of the master. However you look at it, the finished discs cannot properly be described as products obtained directly by means of the patented processes.’

The Court of Appeal accepted that the obverse situation was one in which the intermediate product had lost its identity and had become something else. The products of many, if not most, intermediate processes in chemical cases will suffer that fate. I take it that prima facie the phrase ‘directly obtained by means of the process’ means ‘the immediate product of the process’, or, where the patented process is an intermediate stage in the manufacture of some ultimate product, that product, but only if the product of the intermediate process still retains its identity. In most cases, the assessment will be a matter of fact and degree but not always—Pioneer v Warner was a strike-out case.

What here is the process? So far as the allegation of infringement is concerned, it is defined by the method claims 14 (new claim 5). This opens with the words ‘a method of producing genetically transformed plants…’ and the steps of the method commence with the insertion into the genome of a plant cell a double-stranded recombinant DNA molecule having the prescribed characteristics. This is hardly an everyday operation: it will have been carried out only on the parent of every strain of Round Up Ready soybeans. In fact, it appears to have been done once so far as this action is concerned: the plant identified in Example 3 and Table X of the patent as 13640/40-3 is the parent of all Round Up Ready soybean plants. Dr Padgette gives the history in his witness statement.

The transformation of this plant was many generations ago. Since then, soybeans have been grown by seedsmen or retained by farmers for planting; the plants have been grown and the new beans harvested; and after some generations the harvested beans have been processed into the meal in the Podhale cargo. I accept that all the Round Up Ready soybean plants in Argentina are lineal descendants of this original plant, and I can see how it can be said that this huge mountain of soybean meal (5000 tonnes on the Podhale alone) can be described as the ultimate product of the original transformation of the parent plant. But I cannot see that it can be properly described as the direct product of that transformation, a phrase I would reserve for the original transformed plant. This aspect of the claim must fail.

Monsanto says that the product has retained its essential characteristics. The meal comes from beans produced by a plant which contained the Round Up Ready sequence. It was the sequence that made the invention patentable, and the sequence has survived. Even though the meal comes from beans which are not the beans from the plant which underwent the original transformation, that is enough. I think this has nothing to do with the product of the process at all. It might be extravagant to say that the generation of plants producing the beans from which the Podhale meal was manufactured did not have an atom in common with the original transformed plant, but it must be close to the truth. I think that Monsanto’s argument confuses the informational content of what passed between the generations (the Round Up Ready genomic sequence) with the product, which is just soybean meal with no special intrinsic characteristics from one of the generations of plants. Put another way, it is difficult to see how anything has survived into the meal if the sequence has not. It cannot be told apart from non-Round Up Ready meal unless it contains traces of the gene, in which case other claims are relevant. What has not survived is the original transformed plant. I should add that I think it is dangerous to talk of reproductive material having in some way passed between the generations. While no doubt some reproductive material does pass between the first and second generations, the same material does not pass further. Copies pass thereafter.

A considerable case was mounted by Cargill in support of the contention that Monsanto had acquiesced in Argentinean soybean production using unlicensed Round Up Ready plants. This plea was raised against the background that for reasons not fully explained Monsanto has no enforceable rights in Argentina in respect of Round Up Ready soya. It did not survive the evidence given by Mr Amado, who was a palpably honest and forthright witness. He could not identify anything which Monsanto had done or said from early 2004 which suggested to Cargill that it would be able to export the produce of the 2005 harvest to Europe without complaint. In respect of the cargo loaded on board the Podhale, therefore, no representation was made by Monsanto or relied on by Cargill to its detriment and this plea must fail.

Before turning to the individual claims of the patent, it is necessary to identify what exactly is in the Podhale meal. There are six experiments conducted by Monsanto. Cargill performed no experiments, so restricting themselves to criticisms of the manner in which the experiments were performed by Monsanto and of their intrinsic probative value. This was very unsatisfactory because many of the criticisms made of Monsanto’s experimental technique and results could have been demonstrated experimentally, but were not. So far as the Southern blot and Southern blot/PCR experiments performed by Monsanto were concerned, Dr Isaac had been provided with the restriction digests and the template DNA, in quantities suitable for multiple repeats.

Not all meal shipped by Cargill in Argentina is processed by them. Accordingly the product and process description they provided was described as ‘generic’ and does not call for précis.

‘The process is multi-step, consisting of:

1.

Harvested soybeans are collected from a variety of farmers, combined, cleaned and air-dried at about 180ºF (about 82ºC) to about 9.5-10% moisture;

2.

The beans are then placed in tempering bins for 24-72 hours;

3.

The beans are cracked on rollers to remove the hulls (typically they are cracked to 1/4s and 1/8s of whole bean sizes;

4.

The hulls are separated by aspiration using a flow of air;

5.

The cracked beans are heated at 160ºF (about 71ºC) for about 20-30 minutes;

6.

The resulting material is flaked between rollers to prepare pieces 0.012˝ to 0.16˝ thick;

7.

The flakes are then often put through an expander to form “collets”—this process involves putting the flakes under mechanical and steam pressure, which when released results in the formation of a porous, food like substance which remains on the extractor for a greater period of time;

8.

The collets are subjected to a solvent extraction process (commercial hexane or isohexane) to remove the soy oil;

9.

The defatted soya bean flakes and/or collets are then processed in the desolventizor toaster where direct and indirect steam is used to drive off the solvent. This stage lasts about 30 minutes and the material will reach temperatures of 212ºF (100ºC) for 20 minutes;

10.

The resulting meal is then dried and cooled;

11.

The meal is then ground in particles of either 2mm or 4mm, or pelleted for transport.’

The claims are essentially to DNA sequences. Cargill say that no single– or double–stranded DNA is going to survive this treatment but will be degraded. The material will not infringe.

The five experiments were as follows.

This merely used a commercially available ‘dipstick’ sold to enable users to discover whether their meal came from RuR plants. The dipstick is dipped into a sample of meal which has been blended into water. There are two monoclonal antibodies in the dipstick, which are intended to bind to particular regions (epitopes) on the RuR EPSPS enzyme. They gave a positive result. They do not show that the enzyme molecules were whole (one epitope might have been on one fragment, and the other on another fragment), but the dipstick does show that the meal came, at least in part, from RuR plants and that is accepted.

There are restriction sites for the restriction enzymes HindIII and DraI close to the ends of the RuR insert in the plant genome. The idea is to digest the meal with these enzymes and to run a sample from the digested meal, a sample from undisputed digested RuR beans (positive control) and from digested non-RuR beans (negative control). The DNA samples from each source move side-by-side along lanes through agarose gel under the influence of an electric field. There will thus be lanes for the sample, the positive control and the negative control, and a lane is also provided for a mixture (commercially available) containing appropriate graduated molecular weight samples. Once the samples have moved far enough, the samples are blotted onto a nitrocellulose sheet, where they are immobilised. A DNA probe to the sequence of interest and including radioactive atoms binds to the corresponding immobilised sequences on the nitrocellulose sheet and the location of the bound probe molecules is revealed by making an autoradiograph on photographic film.

Monsanto worked this experiment up and not only the results produced for the notice of experiments and for the repeat were considered at trial, but also the work-ups. The results were visible to the untutored eye (Monsanto Notice Annex 9 and Repeat Report Annexes 3 and 4). There is a band of DNA 3.1 kilobases (kb) long which hybridised with the probe. There is not much of it (the trace is faint) but it shows as a distinctly darker band over a pretty continuous background which it is said is to be interpreted as fragments of the sequence also hybridising to the probe. It appears (again to the untutored eye) that the distinct band does not represent the highest molecular weight present in the sample lane. There is no band in the control lane in the Notice. There is a band to my eye comparable in intensity at 5.1 kb in the negative control lane (1) in the long exposure repeat at Repeat report Annex 4. In the positive control lane, the intensity of the bands is overwhelming.

What I may call the first-order inference from these basic facts as I understand it is as follows. The presence of the band in the sample lane at 3.1 kb suggests that from the soybean genetic material—undoubtedly much damaged by processing in manufacture of the meal—there survives enough double-stranded DNA to produce a 3.1kb sample on digestion with the two restriction enzymes which in the blotting process responds to a probe consisting of the whole RuR sequence. It is therefore the right length and the right sequence.

Three approaches to evidence of this sort is possible. The first is to show affirmatively that it is possible to produce a 3.1kb sequence from meal made from non-transgenic soya, which hybridises to the probe used. The purpose of the negative control lane is to show that no such sequence is produced from non-transgenic soya. Cargill did no such experiment, but did say that at least one gel from the work-up experiments performed by Monsanto indicated the possibility that such a fragment might be produced from non-transgenic soya. The second approach is to demonstrate that the experiment is not probative of the fact sought to be established. This shades off into the third approach, which is to explain the result away by employing either detailed criticism of technique, defects of overall design, possible misinterpretation of results and so on as the case requires.

On the whole, I think Cargill preferred the third approach. Two broad possible explanations were advanced by Dr Isaac for the presence of the 3.1kb band in the Southern blot although it was absent ex hypothesi from the meal. These can be called the “non-specific binding” and the “hairy DNA” theories. The first is more important from the point of view of Cargill’s case, it being said that the probe and conditions used by Monsanto were non-specific. In these circumstances, it is submitted that Monsanto has failed to prove that any relevant DNA is present in the meal.

An evidential point is made by Cargill in this connection. It is said to follow from the speech of Lord Reid in Van der Lely v Bamfords [1963] RPC 61 at 71 that where there was a complete clash of evidence between two competent experts, the court could not hold that the legal onus had been discharged if there was no reason to prefer the evidence of one to that of the other. In Van der Lely, the problem arose in respect of the proper interpretation of a photograph of a hay-rake: what did the photograph show? Lord Reid (Footnote: ²) said this:

‘There is no doubt that, where the matter alleged to amount to anticipation consists of written description, the interpretation of that description is, like the interpretation of any document, a question for the court assisted where necessary by evidence regarding the meaning of technical language. It was argued that the same applied to a photograph. I do not think so. Lawyers are expected to be experts in the use of the English language, but we are not experts in the reading or interpretation of photographs. The question is what the eye of the man with appropriate engineering skill and experience would see in the photograph, and that appears to me to be a matter for evidence. Where the evidence is contradictory, the judge must decide. But the judge ought not, in my opinion, to attempt to read or construe the photograph himself; he looks at the photograph in determining which of the explanations given by the witnesses appears to be most worthy of acceptance.’

The Southern blot is only analogous to a photograph at a level of abstraction so high as to be unhelpful. It is an experimental result, the big variables in which are (1) the distance covered by the DNA fragments in the agarose gel under electrophoresis (2) the extent to which and the specificity with which those fragments hybridise to the probe used and (3) the intensity of bands produced by the decaying radionuclide on the film. Such things are capable of rational exposition, as is the quality of the work done, the uncertainties attached to the technique generally and in the particular case and the proper interpretation of the results obtained. One can see the results. The problem is the meaning which should properly be attributed to those results. I think that the difference between interpreting a Southern blot on the one hand and trying to decide whether a photograph shows a rake whose rotors are rotated by contact with the ground or by invisible gearing is so great that further comment is not called for. In all cases, the expert witnesses are there to provide the judge so far as possible with the eyes of the skilled person. Whatever may have been the position when the case involved a hayrake, in the modern state of things this task is really no different whether obscure photographs or sentences which are incomprehensible are under examination. Let me take an example from the textbook which I used with the agreement of the parties for the purposes of the action (Footnote: ³) to learn about kinetics:

‘Chemical bonds are formed by electrons, and the rearrangement or breakage of bonds requires the migration of electrons. In broad terms, reactive chemical groups can be said to function either as electrophiles or as nucleophiles. Electrophiles are electron-deficient substances that react with electron-rich substances; nucleophiles are electron-rich substances that react with electron deficient substances. The task of a catalyst often is to make a potentially reactive group more reactive by increasing its intrinsic electrophilic or nucleophilic character. In may cases the simplest way to do this is to add or remove a proton.’

With all respect, one cannot get much meaning even out of a very elementary textual passage like this one just with instruction as to the meaning of the technical language. The result would be worse than a translation from a language of which one knew nothing and had no guidance (Finnish, say) produced with nothing more than the aid of a dictionary. The experts must explain to the judge what the judge is seeing in the text or the photograph, and part of that process is to give the judge enough material and confidence to decide between the views expressed by them when they differ. I do not think that it is practically possible for me to decide which of the various views expressed by the witnesses I prefer without being in a position to assess how well their evidence chimes in with what my unskilled eye sees on the page and what I understand of the scientific principles underlying the experimental procedure of which the autoradiographs and the photographs of the gels (if any) are the record.

It should be very rare in modern patent litigation that the court is left with insufficient instruction to make up its mind as to what an experiment shows. This may be no more than saying that there are few cases in which there is no reason to prefer the evidence of one expert over another. This is not such a case, because Cargill performed no experiments and the question for decision, put crudely, is whether they have satisfied me on the balance of probabilities that Monsanto obtained no experimental results probative of any relevant fact.

The objection is that the probe used, which is accepted to be the whole RuR sequence, is not sufficiently specific under the conditions of the blot. The photograph alleged to show this most clearly is not one in the experimental results, but one from the work-ups carried out by Monsanto. It is page 26 of Volume Q, but is exhibited also to Dr Isaac’s Report at PG-1 page 9 (hereafter ‘page 26’). In the left-hand lane, which is the negative control consisting of material extracted from non-transgenic seed, bands are visible . This must mean that there is a degree of hybridisation with non-RuR material. Moreover the probe is not specific to the whole sequence, since it has hybridised to other material in the positive control and a great deal of comparatively short material in the Podhale meal lane. So it is non-specific, and the material of length ~3.1kb in length to which it has hybridised in the Podhale lane cannot with confidence be said to possess the RuR sequence.

While Dr Isaac accepts that the band in the Podhale lane is, indeed, the right length, he suggests that the experiment is, for the foregoing basic reason, valueless. He provides an explanation for the 3.1kb band, which was called the hairy DNA hypothesis. But Monsanto observed that if Dr Isaac is correct, there is a band missing from page 26. This is the band that may be just visible in the negative control lane (it is not visible on any other blot) and represents on this hypothesis a DNA fragment 3.1kb long present in both the transgenic seed (lane 5) and the Podhale meal (lane 3). It should be present in very large quantities in the non-transgenic seed (lane 1). It ought to be very intense because this DNA is not degraded and there is no doubt that the Podhale meal is very degraded. So, where is it? Dr Isaac proposed a further explanation: the digestion of the material for the control lane had been incomplete, meaning that larger fragments have been left. This does not include the two bands at ~10kb, because these are agreed to be irrelevant for present purposes and to be fully digested.

The missing band was equally not present in the repeat experiments, although it is clear in that experiment that there is much more DNA in the negative control lane with a length above 3kb than there is in the Podhale sample. So, there should have been a band (transcript 1188-9):

Partial digestion cannot be an answer to this. Because the restriction enzymes will operate simultaneously, there will never be a sample in which one has operated but the other has not. There will always be some completely digested material. So there should always be some completely digested material, and, if the Podhale band is an artefact some band should be visible in the negative control at 3.1kb.

I find that there is something in the Podhale lane at about 3.1kb not present in the negative control.

So far as the other background bands are concerned, Monsanto says that they are all visible in both the positive and negative control lanes. Accordingly whatever they are, they are irrelevant, and what can be identified is a point common to the positive and Podhale lanes which has no correspondence in the negative lane.

What is the something which I find to be present? Is it the RuR sequence, or is it something else? In any event it plainly must be capable of hybridising to the probe, and there is no doubt that the RuR sequence will do this, and that the probe does possess some degree of specificity. Against its being the RuR sequence is the very severe treatment the meal receives during its processing.

Dr Isaac suggests that in fact the ~3.1kb band is an artefact. While the something is capable of hybridising to the RuR probe, it is made up of single-stranded fragments less than 3.1 kb long which have annealed (joined together) as complementary strands. A result of this process will be that non-matching sequences will form loops or just not anneal, so leaving ‘hairs’ consisting of parts of single-strand molecules sticking out from the main double strand. Obviously this is the explanation if the Podhale meal contains no molecules >~3.1kb long. But it does. The gel obtained after electrophoresis at the repeat experiments and stained with ethidium bromide (bundle Q page 2) shows all the DNA present, of whatever molecular weight. There is plainly DNA in the Podhale sample well over 3.1kb.

This led to a further explanation which I did not find satisfactory. Dr Isaac suggested that the DNA apparently present in the negative control lane on the gel was in fact low molecular weight material that had ‘leached out’ of the sample well at the top of the gel after a delay caused by not being perfectly in solution but entangled in a high molecular weight complex. Obviously this would be a small effect, or all the bands would be obscured. I found this to be a rather ad hoc explanation which emerged to negative a perfectly straightforward point.

Professor Lichtenstein, for Monsanto, calculated the likely rate of re-annealing of the fragmentary strands during the extraction of DNA from the Podhale meal, and obtained a very low figure. Indeed, one can see that if only two molecules need to come together, that is a possibility; but the chances of 4 appropriate molecules meeting up sequentially would be very low. Professor Lichtenstein carried out a calculation which was not effectively challenged suggesting that the rate would be too low to produce detectable quantities of DNA, and Dr Isaac said that the limits would have to be detected experimentally: this was one of a number of points where an experiment successfully conducted by Cargill would have counted heavily.

Finally, it must be remembered that the re-annealed sequence is going to have a considerable range of lengths in consequence of the loops and ‘hairs’. It must also be remembered that the mechanism would result in the production of larger structures (albeit slowly). On the whole I thought that the straightforward explanation—that the presence of some DNA with the RuR sequence in the Podhale meal was more likely than not on the basis of this experiment, and I considered that my view was reinforced by the final experiment, in which the presence of such material was amplified by the use of the polymerase chain reaction (‘PCR’).

The polymerase chain reaction will amplify a complete ‘template sequence’ using a supply of primers which correspond to the complementary ends of the sequence. There is no doubt that PCR produced the whole of the RuR gene using the DNA extracted from the Podhale meal. That DNA was therefore present in the starting material for PCR. There are two explanations other than the straightforward one: contamination (only one molecule is required) and creation of the template on the fly at the beginning of the PCR sequence, by means of a reaction called splice overlap extension. That this may happen is a consequence of starting with an extract which must contain fragments of the sequence of interest.

Contamination is not alleged. Splice overlap extension seems never to have been used to create a gene as large as the present one, and Professor Lichtenstein calculated the relevant probabilities given the size of the soybean genome and the amount of material extracted from it for a favourable SOE reaction as opposed to an immediate annealing with one of the primers which were present in a very large chemical excess. This evidence was not effectively challenged or answered and I find that the chances of this reaction creating a template molecule which would then be amplified by PCR to be very small. While Cargill are right when they say that quantitative PCR would have solved all problems, I observe that this is a straightforward experiment of a kind carried out by Dr Isaac’s company. It would have been nice to see the effect taking place, preferably with Podhale meal, once ordinary PCR strongly suggested that the intact sequence was present in the meal.

Monsanto also relied on microscopic examination of stained samples of the Podhale meal, the stains selected so as to show, it was said, the presence of double-stranded DNA. The stain is called acridine orange, and the manner in which the results were presented contained an error. Paragraph 6.38 of the Notice does not describe any of the results achieved, although this is not apparent at a first reading.

I can find no technical answer to this experiment, once the Notice has been deciphered. Double stranded DNA is present, and the only question is whether it is present in the meal, and whether the RuR gene is double-stranded. This depends in its turn on whether any of the DNA in the meal has been denatured, which is a different matter from its disruption and fragmentation. I accept that it is likely that if some of the DNA in the meal is double-stranded, there can be no reason to suppose that the RuR gene, if present, will not be.

Finally, I should refer to the literature, such as it is, on the effect of processing on recombinant DNA in soybean meal. I have no doubt on the basis of Dr Isaac’s exhibit “Report for Food Standards Agency Project Code Number: FOI004” by Phipps et al. of the Centre for Dairy Research and of the ADAS Nutritional Sciences Research Unit that it is possible to process soybean meal so as to degrade DNA very substantially, leaving fragments containing a few hundred bases. It seems on the basis of Smith et al (2000) that much larger fragments, potentially large enough to contain a functional EPSPS gene, are also capable of surviving processing. As one might expect, it all depends on the conditions. It is unfortunate that Cargill were very sparing in the information which they provided in respect of their own process, although reticence in respect of meal purchased from other processors in Argentina is understandable. So I cannot exclude the presence of a complete RuR gene on the basis of the processing.

Finally I come to the evidential position. The Southern blot and PCR experiments have prima facie demonstrated the presence of the RuR gene in the meal. Samples of the materials have been provided, and Cargill have carried out no experiments tending to demonstrate the absence of the gene, or that the experiments carried out by Monsanto were unreliable. The whole has been based on Dr Isaac’s criticisms and comments. It was established that the ‘hairy DNA’ hypothesis could be experimentally tested, since digestion of this material with mung bean nuclease would cut hairy strands in the gaps (nicks) between the ends of the annealed fragments. This would destroy the 3.1kb band if it comprised the result of annealing a number of fragments. So too would be to carry out the electrophoresis on a gel which denatured the DNA. On a denaturing gel, the ‘hairy DNA’ would be reduced to its fragments and run accordingly.

Dr Isaac did not explain the PCR results on the basis of contamination, but on the basis that the template was not present in the DNA extracted from the meal. Given that Dr Isaac is himself an expert in analysis, and his company provides analytical services, he also could have performed real-time quantitative PCR to demonstrate that the template did not exist in the original sample meal but was assembled from the DNA fragments in the early cycles of the PCR. This seems to me to be the conclusive way of showing that the template was an artefact and that the Monsanto experiment lacked any value.

On the whole, I find that although the Monsanto experiments are open to criticism, I was not provided with any hard results by Cargill which threw sufficient doubt upon the results that appear to have been achieved. I find as a fact that there was present in the Podhale meal some genomic DNA which included the RuR EPSPS gene, and that some or all of that DNA was double-stranded.

I can now turn to the issues of construction that arise in relation to the allegation of infringement. These issues are as follows. First, is the RuR gene which I have found to be present in the Podhale meal ‘isolated’ (amended claims 1-3). Second, what is ‘a Class II EPSPS enzyme’ and does the RuR sequence in the meal encode for it. Third, does the peptide which it encodes have the kinetic characteristics required by the claim.

This is a surprising word to use when the ultimate destiny of the gene is to be incorporated in the genome of a plant to be inherited by future generations upon which it will confer Round Up resistance. It is quite inappropriate to refer to a sequence present in a genome, even if that sequence is exogenous.

The specification itself uses the word in a straightforward sense. In [0005] it is used to mean isolating the gene from particular sources. In [0009] it is used in the same general sense to refer to isolation of Class II genes from particular bacteria. In [0010] it is used to refer to the obtaining of new Class II genes using a CP4 probe, with a view to expressing the gene so isolated in plants using known technology. In the same paragraph it is used to refer to bacteria isolated from soil, that is, identified and separated from all the other material. The word is used in [0023] really just to mean ‘obtained’:

‘Promoters which are known or are found to cause transcription of DNA in plant cells can be used in the present invention. Such promoters may be obtained from a variety of sources such as plants and plant DNA viruses and include, but are not limited to, the CaMV35S and FMV35S promoters and promoters isolated from plant genes such as ssRUBISCO genes.’

It is used in the same sense of ‘obtained’ in [0028] and again in [0040] where it is used to describe the obtaining of cosmid transformants containing the candidate genes. Generally the word is used to mean identification of the gene and its separation: see [0066] and [0068]. In the same way, chloroplasts are spoken of as having been isolated from lettuce by centrifugation.

I have found it very difficult to work out the sense of this rather simple word in its context in the claim. It would not be unfair to say that the task confronting Monsanto on the issue of infringement was to see whether they could isolate an RuR gene from the Podhale meal, in which case it might be thought that in the Podhale meal, and in the plants from which it was derived, the gene was not isolated. The same problem afflicted Professor Lichtenstein. (‘Isolated’ has a technical aspect and I thought that the experts were helpful on this issue.) He had great difficulty under cross-examination in explaining the word, tending (I think) to equate it to ‘exogenous’ in the sense of ‘originating in another organism’ or ‘no longer in its natural context’. Professor Leaver gave it a very restrictive meaning, essentially saying that what it conveyed was that the sequence had been separated out as a fragment for further cloning and amplification in a plasmid DNA.

When one considers the structure of the specification itself, there is a clue. The unamended claims start with the isolated sequence encoding the class II enzyme, and proceed (claim 6) to a non-isolated sequence encoding particular enzymes, then a series of claims to recombinant double-stranded molecules including a promoter and a 3 end, followed by a method of transforming plants using such a double-stranded molecule (claim 14). Then claim 20 claims the transformed plant cell, and claim 24 the plant. Weeds are controlled in a field full of such plants according to the method of claim 28. This follows the structure of the descriptive part of the specification itself.

What is striking about these claims is that down to claim 14 (the method of transforming the plant) they all relate essentially to the laboratory work. In principle, an infringer need only ever infringe these claims once, but immediately he transforms a single plant all its progeny for the future will infringe. As I have said, the Round Up Ready soya is the progeny of a single transformed plant. It makes sense in this context to have claims relating to each of the stages in transforming the plant, starting with the isolated sequence and proceeding through the sequence appropriately topped and tailed to transforming a plant using that molecule.

I conclude that in the claims the word ‘isolated’ has precisely the meaning that Professor Leaver gave to it, that is, ‘separated from other molecular species in the form of a purified DNA fragment’ for the purposes that he identifies. In its context in this specification, I conclude that this is what the claim is talking about. It is not talking about genomic DNA, of which it is the antithesis, and it is not talking about the DNA of the progeny of a plant or plants transformed using a plasmid incorporating DNA having that sequence.

The difference between the parties is simply stated. Monsanto say that in its context in the claim it is plain that this phrase is defined by the claim itself, the key words being ‘said enzyme being’:

‘encoding a Class II EPSPS enzyme,

said enzyme being an EPSPS enzyme having a K_m for phosphoenolpyruvate (PEP) between 1-150 µM and a K_i(glyphosate)/K_m(PEP) ratio between 3-500

which enzyme is capable of reacting with antibodies raised against a Class II EPSPS enzyme selected from the group consisting of the enzymes of SEQ ID NO:3 and SEO ID NO:5.’

Cargill submit that the use of the phrase Class II EPSPS enzyme imports the whole of paragraph [0009], or at least the second sentence, which I quote again for convenience:

‘Class II EPSPS enzymes can be readily distinguished from Class I EPSPS’s by their inability to react with polyclonal antibodies prepared from Class I EPSPS enzymes under conditions where other Class I EPSPS enzymes would readily react with the Class I antibodies.’

This, it is said, is the only passage in the specification which sets out the features of Class II enzymes, and mere compliance with the words of the claim will not necessarily imply membership of the class at all. What in [0009] is a positive requirement (‘inability to react with polyclonal antibodies raised to Class I EPSPS enzymes’) has become in the claim ‘capable of reacting with antibodies raised against a [single] Class II enzyme’. The two requirements are not the same unless it can be stated that an ability to react with antibodies raised against one of the Class II enzymes specified necessarily implies in all cases an inability to react with Class I antibodies. There is nothing in the specification to suggest that it does.

To adopt the construction of the claim suggested by Monsanto is to deprive the ‘Class II’ concept of most of its content. It is reduced merely to a way of describing any enzyme with the necessary kinetic qualities which will react with antibodies raised to one or other of the specified enzymes, a criterion which is, so far as the specification is concerned, entirely mysterious as no such experiment is described and no results are compared.

Cargill submit that Monsanto and the Technical Board of Appeals in T 179/01 accepted that in its context in the claim the phrase ‘Class II EPSPS enzyme’ in the claim imported the requirement of paragraph [0009], that is, no cross-reactivity with class I antibodies.

I accept that the EPO accepted the construction for which Cargill contend, and disposed of an allegation of anticipation on the basis of it. It seems clear that this was Monsanto’s preferred interpretation in the EPO proceedings. Nonetheless it is my view that Monsanto may argue in this court for a different interpretation, although this case is not one of ‘file wrapper estoppel’ or a reference to a submission to the EPO by a patentee’s representative in the course of the prosecution. As Jacob J observed in Bristol Myers Squibb v Baker Norton [1999] RPC 274, the fact that claim construction is governed by Article 69 EPC and its Protocol gives it something of an autonomous aspect, and it is plain from Kirin-Amgen v Hoechst [2004] UKHL 46, [2005] RPC 9, page 169 that reference to the EPO file is discouraged here and elsewhere for a number of reasons (see [35]) one of which—that life is too short—I am very conscious of. However, here we are considering an interpretation arrived at in contested proceedings by a Board of Appeals of the EPO which should be treated with respect. I think one needs a good reason to depart from an interpretation placed on a claim by a TBA in contested proceedings and which forms part of one of the grounds of decision.

In my judgment, the logic of the definition requires a construction following that of the TBA. As I have said, if I treat the claim alone as defining ‘Class II’ that term is deprived of its substance. It is just a Class consisting of any material having the defined kinetic characteristics, which in any event is all that is required for glyphosate resistance. To say that a class is defined by stating that its members react with antibodies raised to other members of the same class says nothing about the limits to the class, unless it is also possible to exclude members on the basis of their membership of that other Class, i.e. Class I. If this is not possible, there is potential overlap with Class I. In any event, it seems to me that the teaching of the specification is clearly otherwise:

‘[0034] When crude extracts of CP4 and LBAA bacteria (50 µg protein) were probed using rabbit anti-EPSPS antibody (Padgette et al. 1987) to the Petunia EPSPS protein in a Western analysis, no positive signal could be detected, even with extended exposure times (Protein A-¹²⁵I development system) and under conditions where the control EPSPS (Petunia EPSPS, 20 ng: a Class I EPSPS) was readily detected. The presence of EPSPS activity in these extracts was confirmed by enzyme assay. This surprising result, indicating a lack of similarity between the EPSPS’s from these bacterial isolates and those previously studied, coupled with the combination of a low K_m for PEP and a high K_i for glyphosate, illustrates that these new EPSPS enzymes are different from known EPSPS enzymes (now referred to as Class I EPSPS).’

The phrase ‘coupled with’ seems to me to exclude any test for membership of Class II that depends only on possession of the desired kinetic characteristics. Oddly, this passage is the only description of any immunological experiment showing a reaction of CP4 to Class I antibodies, as to which the documentation in disclosure is not satisfactory. Exhibit X3a is a much enlarged Western blot (analogous to a Southern blot in which the protein(s) of interest are subjected to electrophoresis and are probed with radiolabelled antibodies) showing a reaction in the crude extract (lane labelled CP4) but no reaction in the CP4 standard lane.

Given these various factors, I consider that the correct approach to this claim is to take it that a Class II enzyme is one that does not react with antibodies raised to a Class I enzyme. This has not been demonstrated with the material expressed by the RuR gene.

Reaction with antibodies raised to SEQ ID:NO 3, ie. CP4, do not give a problem. While the only experiments were performed with two monoclonal antibodies rather than the polyclonal antibodies called for by the claim, the only difference between the two enzymes lies in the serine to leucine point mutation at the 2 position, or S2L. Nobody suggested that this region formed any part of an epitope (a region of the molecule to which the immune system will respond in producing antibodies). Professor Leaver regarded this as definite, and Dr Isaac considered it a fair assumption.

The kinetic qualities of the RuR EPSPS have not, it seems, been measured, it being assumed that they will be very similar to CP4 EPSPS. Again, it is assumed that the S2L mutation will have a negligible effect, the N-terminus of the protein being far removed from its active site. On the whole, I think this can be accepted.

The allegation of infringement of claim 1, and all claims requiring an isolated sequence, accordingly fails. All claims requiring a sequence to be capable of encoding a Class II enzyme fail because it has not been shown that RuR EPSPS does not react with antibodies raised to a Class I enzyme. Were there otherwise infringement, I would not regard the S2L mutation as preventing infringement, even though the relevant DNA sequence is nowhere claimed, because I am satisfied on the evidence both that the mutation makes no difference to the performance of the enzyme, which is the ultimate target of the invention and that if that were not obvious to a skilled man an experiment would immediately demonstrate it. This point does not, however, arise.

I should also deal with a de minimis point. The DNA present in the meal, such as it is, is entirely irrelevant to the meal as an animal feedstuff, is present in small, variable, quantities and may not be present at all if processing conditions are changed. It is not in any serious sense genetic material. It is just the remains of the material which was in the soybeans from which the meal was extracted. This, it seems to me, is irrelevant. It may raise a question on damages, that there is no causative relationship between acts of infringement, as opposed to acts which are not infringing by English law, and the loss suffered by Monsanto, but this was not argued. There is, generally, no authority in favour of trace quantities of infringing material being held not to infringe, and some authority against it. In any event, I had no proper estimate of the quantity of DNA in the Podhale meal that survived: Professor Lichtenstein put it at 5% of the DNA in cross-examination but the issue was not properly pleaded.

Invalidity is alleged on the basis of anticipation by a paper entitled ‘The organisation and nucleotide sequence of the Bacillus subtilis hisH, tyrA and aroA genes’ Gene 49 (1986) 147-152 (Henner), and by the availability in nature of the bacterium PG2982 with its description in a paper by Moore, ‘Isolation of a Pseudomonas sp. which utilizes the phosphonate herbicide glyphosate’ in Applied and Environmental Microbiology 46 (2): 316-320 (1983).

Obviousness is alleged on the basis of Henner and Moore, EP-A-0218571 (‘Shah’) published in 1987, and ‘Glyphosate Sensitivity of [EPSPS] from Bacillus subtilis Depends upon State of Activation Induced by Monovalent Cations’ published in Archives of Biochemistry and Biophysics 256 (1): 325-334 (1987) (‘Fischer’).

Allegations of insufficiency of two types are made. First, it is said that there is no useful teaching in respect of any gene encoding for any sequence other than CP4 and LBAA. The claim is too wide. It is said that enzymes reacting to antibodies raised to SEQ ID No: 3 and SEQ ID NO:5, the only limitation in claims 4 and 5, may or may not exhibit glyphosate resistance, and it is said that the kinetic parameters do not effectively limit claims 1-3 to glyphosate-resistant EPSPS either. This is an allegation of failure to enable the claims across their width, or Biogen insufficiency.

It is also said that there is no useful definition of ‘Class II EPSPS’.

Finally, it is contended that the specification is wholly insufficient in its description of the assay to be used for the immunological test. Particular areas of doubt or uncertainty are alleged to be whether the antibodies are monoclonal or polyclonal; what animal or animals are to be used to raise the antibodies. Nor is the type of assay specified. If the definition of Class II in paragraph [0009] is referred to as stating that polyclonal antibodies are to be used, it is said that there is no guidance as to which ‘Class I’ EPSPS is to be used to raise the antibodies, the animal(s) are not described and the assay is not specified. Essentially the complaint is that both the ‘Class I’ and the ‘Class II’ immunological assays are insufficiently described. This is what can be called classic insufficiency.

The so-called Syngenta Experiments form an important element of Cargill’s case, both on anticipation by Henner and on the insufficiency of the immunological test of claim 1. Their nature is comprehensively described and a full chronology given in the judgment of Kitchin J of 11 May 2007, a month before the trial, when he refused to admit the experiments on an application by Cargill for that purpose and adjourned Cargill’s application to be disposed of at the trial. After reviewing the facts, and considering in particular the fact that Monsanto had seen the experiments before in the EPO, he said this:

Among the factors which Kitchin J had to consider were the lateness of the application, Cargill’s apparently deliberate disregard for the procedural rules relating to experiments and the evidence of prejudice as it then appeared. He left the application with the parties clearly understanding that if Monsanto could point to real prejudice arising from admission of the experiments, the experiments should not be admitted. This is in a sense a partial exercise of the discretion to admit the experiments and because the parties both regarded Kitchin J’s findings as potentially decisive of the issues apart from prejudice, I should not normally revisit them. But Monsanto says that in fact Kitchin J’s approach, hostile to Cargill’s application as it was, proceeded upon the false basis that the evidence to be given by Dr Hawkes, would be that which appeared in his evidence in chief.

The first point is a technical one. The Syngenta experiments include an experiment to test the enzyme of interest against antibodies raised to CP4 EPSPS, that is, to the enzyme of SEQ ID NO:3. The enzyme of interest is B subtilis EPSPS. There is no evidence as to how the B subtilis EPSPS was expressed, and it appeared from Dr Hawkes’s witness statement that the CP4 EPSPS had been expressed in E coli, which itself contains a gene expressing E coli EPSPS, which so far as the patent is concerned is a Class I enzyme (see page 2 lines 23-6). A signal in the Western Blot which came, not from antibodies to CP4 EPSPS, but from anti-E coli EPSPS is not within the claim. How then to remove the possible E coli EPSPS antibodies and with them the possibility of a false positive? The technique used is called an affinity column, and it works by immobilising pure CP4 EPSPS on the column, and running the possibly contaminated serum over the column. The idea is that the anti-CP4 will bind to the CP4, the remaining antibodies passing through. But to do this, it is necessary to have some pure CP4 EPSPS, and the only way of achieving this, if one is expressing in E coli, is to use a strain of E coli which does not express native EPSPS, because it lacks the necessary gene. Such a strain is called , i.e. lacking the gene . This requires an E coli mutant in which the gene is not present, not a mutant in which the gene is present but expresses part of EPSPS, which may still contain regions that react with the anti-E coli. The purification of the anti-CP4 material is described in Annex 3 of the Syngenta experiments.

To cut a long story short, Professor Lichtenstein said the CP4 EPSPS used in the affinity column would have to be expressed in an deletion mutant E coli, and it was confirmed that this had happened. Professor Leaver was cross-examined on that basis. When Dr Hawkes gave evidence, it turned out that the gene used to express CP4 EPSPS in E coli was not a deletion mutant but an insertion mutant, and he did not know how much of the gene had been excluded by the mutation (the insertion of a tetracycline resistance gene at some point). He could not say with certainty that the insertion had prevented the expression of a length of protein present in EPSPS which would, of course, be antigenic.

Other material in respect of the experiments was certainly in existence and had not been produced. It is obvious that the blots reproduced in the experimental results produced by Dr Hawkes have been electronically processed. The original blots are probably available from the records, on which Dr Hawkes was not authorised by his employers, Syngenta, to expend his time. Nor had he done the immunological experiments, having himself done protein expression and purification at Jealott’s Hill in Berkshire, while the immunological work, producing the antibodies in the selected animals, running the gels and producing the blots was done by colleagues at Alderley Park in Cheshire. Dr Hawkes did not witness these experiments, although he had full confidence in the scientists who had performed them.

It is also a matter for comment that while Professor Lichtenstein has said that he would prefer to see full notices of experiments and have the opportunity to witness a repeat, his desire to do so has been criticised:

‘Apart from the, frankly desperate, suggestion that Zeneca scientists with substantial experience in running gels could not load the gels properly, Professor Lichtenstein’s evidence is merely that he would have liked to have watched everything to see if he could find fault.’

This may be compared with Dr Isaac’s view expressed in his cross-examination (transcript 1112):

I should also refer to a number of other points. Cargill do not seek to put in evidence all the experimental material upon which Syngenta relied in the EPO. I do not understand why they have effectively omitted all the material about expression, purification and characterisation of the proteins used in the Syngenta work. There is an obvious problem with the description of the bleeds, for example, which needs explanation which Dr Hawkes cannot give.

Procedurally, I think there is prejudice. Professor Lichtenstein does not have samples to take home and experiment on: this is a basic facility which anyone conducting experimentation of this kind would do well to provide. All the actual experimenters do not give evidence, and the data has been ‘cleaned’ for presentation. This fact is not sinister, but procedural fairness requires access to the raw data unless the expense and inconvenience is wholly disproportionate. It would enable Monsanto to carry out the same operation as Cargill carried out with Bundle Q, the results of Monsanto’s work-up experiments in the Southern blot experiments. The significance of the use of the insertion mutation has not been (and likely cannot) be assessed without information relating to the mutation not presently available. Finally the evidence is that the experiment is not repeatable given the present documentation (Footnote: ⁴). On the other hand, if the experimental material is admitted, a strong attack against the patent is provided.

I am in no doubt that there can be experimental evidence which is straightforward, and in respect of which the court will not make orders for repetition and inspection without very good reason. It may well be that it will wait to see good evidence, or a repetition, by the other party that throws doubt upon the result obtained before repetition is ordered. Normally, though, I consider that parties to litigation of this degree of complexity are entitled to inspect a repetition and take samples before the experiment may be relied on against them. While one may entertain the forensic conviction that there can be no answer to a particular experiment, they should have every opportunity of satisfying themselves of that fact. This is not to encourage experimental evidence: it is to ensure that experimental evidence deployed in court has the additional strength which comes from providing the other side with every opportunity to test it.

I am particularly anxious that the regime for managing experimental evidence provided by the rules is not weakened. There is nothing worse than a poorly documented, unrepeated experiment coming into proceedings at a late stage, as anyone who has worked in the EPO will attest. In my view experimental evidence should be kept to a minimum, and that minimum should consist of inspected experiments. Admission of late, uninspected experiments does nobody any favours, unless perhaps to introduce an element of surprise which I am equally anxious should be eliminated so far as possible.

Kitchin J was particularly concerned about the antibody purification step. The actual facts about this step became known only during trial and might be significant—the parties have no time to carry out experiments. He was also concerned generally about the question of contamination with anti-E coli EPSPS. On the information available this risk may well have been excluded, but on such a question as this such a conclusion is not enough. Monsanto ought to be able to perform experiments with a view to demonstrating that the balance of probabilities is not as it presently appears, but they can do no reply experiments in the circumstances of this case. I think this amounts to real prejudice, and I exclude the Syngenta experiments accordingly.

This case raises no new principle, but it provides examples of factual circumstances coming very close to the line. A claim is anticipated by a prior written disclosure if the prior document provides a clear description of, or clear and unmistakeable directions to do or make, something which falls within the claim. In the case of a description of something within the claim, the prior document and the common general knowledge must between them enable the thing to be made without the exercise of ingenuity and without the performance of anything more than routine experimentation and testing. It should be noted that when the case is based on clear and unmistakeable directions contained in the prior document the skilled man may only use his knowledge of the result called for by the claim if that result is, itself, disclosed by the prior document. To this extent, the requirements of sufficiency of description, where the end is known, and of enabling disclosure may differ. An anticipation by prior use must equally be enabling in a relevant sense. See generally SKB v Synthon sub nom. SmithKline Beecham plc’s (Paroxetine methanesulfonate) Patent [2005] UKHL 59, [2006] RPC 10 (page 323).

Obviousness is an enquiry as to the effect on the mind of the uninventive skilled person, possessed of the common general knowledge in the art but of no other matter, of the disclosure alleged to make the invention claimed obvious. The standard analysis followed is that of Windsurfing International v Tabur Marine [1985] RPC 59 (CA):

‘There are, we think, four steps which require to be taken in answering the jury question. The first is to identify the inventive concept embodied in the patent in suit. Thereafter, the court has to assume the mantle of the normally skilled but unimaginative addressee in the art at the priority date and to impute to him what was, at that date, common general knowledge in the art in question. The third step is to identify what, if any, differences exist between the matter [forming part of the state of the art] and the alleged invention. Finally, the court has to ask itself whether, viewed without any knowledge of the alleged invention, those differences constitute steps which would have been obvious to the skilled man or whether they require any degree of invention.’

The Henner paper discloses the sequence of B subtilis EPSPS together with a DNA sequence encoding it (it calls it the gene). It is accepted that this enzyme has the kinetic qualities called for by the claims of the patent. Accordingly, the question is whether this is a Class II enzyme, and whether it reacts to CP4 antibodies as the claim requires. The latter point was sought to be demonstrated by the Syngenta experiments, which I have excluded.

It is not shown that the B subtilis EPSPS does not react with antibodies raised to any class I enzyme. So it is not shown directly that it is a class II enzyme as I consider this term is to be understood in the patent in suit. This leaves the statements contained in Monsanto’s own patent, US 5,627,061 (Barry) which describe B subtilis EPSPS as a class II enzyme (column 3 line 56 – column 4 line 4, column 4 lines 52–7). This disclosure is different from that of the patent in suit so far as it relates to the distinction between Class I on the one hand and Class II on the other, the clear statements as to lack of reactivity with antibodies raised to class I enzymes being modified somewhat:

‘Class II EPSPS enzymes often may be distinguished from Class I EPSPS’s by their inability to react with polyclonal antibodies prepared from Class I EPSPS enzymes under conditions where other Class I EPSPS enzymes would readily react with the Class I antibodies as well as the presence of certain unique regions of amino acid homology which are conserved in Class II EPSP synthases as discussed hereinafter.[my emphasis]’

This passage reveals a certain lack of confidence in the proposition advanced in [0009] of the patent in suit, but that document must be interpreted on its own terms, and without reference to any second thoughts the patentee might have published thereafter.

B subtilis is identified in Barry as an exemplary Class II enzyme. There is a strong inference from this document that it both fails to react to Class I antibodies and reacts with Class II antibodies, but its immunological characteristics are not disclosed, and so it is not possible to determine whether this is a Class II enzyme as I consider that term must be defined in the patent in suit.

Finally, there was some discussion in Professor Leaver’s evidence of his expectation that B subtilis EPSPS would react with anti-CP4. Again, on this issue there was much speculation, including questions to Professor Lichtenstein on the inferences to be drawn from a computer analysis by Monsanto carried out long after the priority date in which a possible epitope possessed by both B subtilis and CP4 was identified and which gave a good score to the possibility of a common epitope at the N terminus of the protein. This is far too thin a basis for coming to the conclusion that B subtilis would, in fact, react with anti-CP4. The fact that I have excluded the Syngenta experiments conceals the fact that properly performed experiments would have resolved these questions. It follows that Cargill have failed to demonstrate anticipation by Henner.

The case of obviousness of the invention in the light of the Henner disclosure can be dealt with here. It is not immediately clear from the text of Henner that it is talking about the EPSPS gene in B subtilis when it describes the gene. There is no teaching as to the kinetic qualities of the EPSPS and no teaching as to its immunological qualities. The case is thus really the same as the case on anticipation: one must assume, I think, that it is obvious to express the sequence set out at length in Henner, and if one does so one will obtain B subtilis EPSPS. The evidence does not show what its qualities would be, and so the case based on it fails.

The title of the paper is ‘Isolation of a Pseudomonas sp which utilises the phosphonate herbicide glyphosate’. The strain in question is PG2982 which, as the patent demonstrates, expresses a class II EPSPS identical to LBAA that possesses the required kinetic qualities (see [0009] and Table 1). Accordingly there are in the paper clear directions to make an isolated strain of bacteria containing DNA encoding an enzyme within the claim.

For the reasons I have already discussed in considering infringement, this bacterium does not contain an ‘isolated’ sequence. The sequence is part of the bacterial genome. Cargill submit that claim 6 is anticipated, since it is not limited to an isolated sequence. Monsanto submit that the bacterial strain of Moore is not made available to the public.

Moore describes the isolation of the bacteria of interest from a stock culture of Pseudomonas aeruginosa ATCC 9027 (the reference indicates it came originally from the leading culture collection, the American Type Culture Collection). Monsanto do not allege that the directions given in Moore are not sufficient to isolate the strain in the manner described, but say that the disclosure of Moore is of a bacterium that can be isolated from Moore’s own Pseudomonas aeruginosa culture, there being no suggestion that it was available elsewhere: and the evidence of Dr Padgette of Monsanto was that it was probable that he had obtained the sample of PG2982 from Dr Moore. This, Monsanto say, is not enough to say that the strain was made available to the public by this publication, which is the only thing relied on.

It seems clear enough that there was nothing secret about PG2982, and the evidence was that Dr Moore supplied it when asked. The disclosure of the document is directed to glyphosate tolerance, although it is plain that it discloses no sequence, and a gene conferring glyphosate tolerance on E coli sequenced by a connected team was not the EPSPS gene. I am of the view that once the existence of the strain is disclosed and once it is clear that Dr Moore had supplied a sample, that is enough to make the microorganism available to the public and to anticipate claim 6.

While the anticipation attack on the basis of Moore is limited to claim 6, a wider obviousness attack is made on the basis of the same document. Moore discloses the existence of PG2982, a glyphosate-resistant organism in fact containing the CP4 sequence. It is disclosed as a microorganism that metabolizes glyphosate. Its culture is described as the ‘glyphosate-utilizing culture’ (page 318, for example) and the next step for the authors is described at the end of the paper:

‘We are interested in elucidating how PG2982 breaks down glyphosate and whether a “phosphonatase” enzyme similar to that characterised in B cereus by [reference omitted] is involved in the degradation. Assays for the degradation of glyphosate and Amph with undialyzed and dialyzed cell-free extracts have been unsuccessful…’

If it was obvious to these authors that the resistance to glyphosate necessary for the bacterium capable of metabolizing it was conferred by the particular qualities of EPSPS, an enzyme having analogues in all plants and bacteria, they weren’t saying. Cargill put the argument like this. It was clearly obvious to investigate [PG2982] (Footnote: ⁵) because of its apparent ability to grow in a glyphosate-rich medium. Whatever the mechanism (neutralisation of glyphosate or glyphosate resistance) it called for investigation. This it is suggested was confirmed by Professor Lichtenstein in cross-examination in the following passages (transcript 555 and 7):

This is a less than ringing endorsement of glyphosate-degrading micro-organisms as the way forward. It was Monsanto’s approach, which was in effect to screen microorganisms which degraded glyphosate for EPSPS which was glyphosate-resistant. Such an open-ended approach is capable of supporting a claim, if (among other things) the field is wide and it is not apparent that the efforts are likely to be crowned with success.

The reason, say Monsanto, that it was not apparent that the efforts would be crowned with success requires a closer look at the common perceptions in the art at the priority date. One starts with the mechanism of inhibition by glyphosate. Glyphosate’s mode of action by competitive inhibition of EPSPS was common general knowledge. Monsanto submit that a glyphosate-resistant EPSPS was, at the priority date, something generally known to be desirable, what is being looked for is an EPSPS in which the K_i and K_m have been ‘decoupled’, and it was not obvious that this could be done.

I do not think that Monsanto give a correct description of the problem confronting the skilled man. In paragraphs [13] ff above I explain the nature of K_m and K_i. These are parameters which serve to classify an EPSPS: they say that it catalyses the reaction at such-and-such a rate while showing such-and-such an affinity with glyphosate. What the skilled man is looking for is an enzyme that can be expressed in a plant and which will bind sufficiently to its substrate PEP to produce enough product that the growth of the plant is not seriously impaired while competitively binding with glyphosate at a rate which will be as low as possible. The range of K_m and K_i/K_m covered by the claim covers effectively everything that is adequately resistant, and is expressly described as a feature of Class II EPSPSs—see [0012] in the patent. It is just a description in numerical terms of ‘sufficient resistance of EPSPS to glyphosate to enable the plant to grow’.

In my view, it must be recognised that resistant microorganisms are not the only targets for the skilled addressee’s search. It is clear that the possibility of over-expression (i.e. increasing the amount of EPSPS expressed by recombinant means), introduction of a glyphosate degradation gene and mutation of the EPSPS gene to obtain mutants with unknown properties were being actively pursued before the priority date. For example, Monsanto published (Kishore & Shah (1988)) a description of over-expression of Petunia EPSPS in transgenic petunia conferring glyphosate resistance. This had disadvantages: the plants grow more slowly in consequence of the accumulation of glyphosate in the undifferentiated cells. Professor Leaver expressed the view, unchallenged, that this meant that the skilled addressee would consider that the most commercially useful manner of conferring glyphosate resistance was likely to be mutation. The company referred to in the proceedings as Calgene pursued mutant Salmonella typhimurium EPSPS. The Ruhr Group pursued mutant Klebsiella pneumoniae (Monsanto were associated with this work) and Monsanto did other work on mutant EPSPSs as well, which is referred to in the pleaded patent specification EP-A-0 218 571 (Shah), published in 1987. This work appears to have led nowhere, with the exception of some of the Calgene work.

The general experience appears to have been that the effect of mutation was to raise K_m, of the specific enzymes, which is not desirable. Professor Leaver’s evidence did rather suggest a hope for victory confronted by technical adversity (transcript 759):

When the Professor talks of decoupling the increase in K_i from the increase in K_m, he is saying that in every case at that stage an improvement in the enzyme’s unwillingness to bind to glyphosate had been accompanied by an increased unwillingness to bind to its substrate. What Professor Leaver is identifying is a hope that such an EPSPS which was unwilling to bind to glyphosate but whose affinity for its substrate might exist, although the mechanism by which glyphosate operates is a competitive one. That is confirmed by another passage upon which Monsanto rely (transcript 802):

Professor Lichtenstein definitely lacked Professor Leaver’s optimism. He said that he would not expect to obtain an EPSPS with the desirable combination of qualities.

Cargill also sought to suggest that it was common general knowledge in the art that a transformed plant had been produced before the priority date in which K_m had been maintained at the low level of the plant in nature but with K_i substantially increased, and so the rate of binding to glyphosate reduced. The paper is by Comai, and the source is Calgene. The publication is in a journal suggested to be obscure: “Vorträge für Pflanzenzüchtung”. The paper appears to report the successful transformation of tomato plants with a mutant bacterial (S typhimurium) EPSPS, and delivery of the enzyme into the chloroplast by a transit peptide. The document is primarily directed to the transit peptide. It is unpleaded. The properties of the S typhimurium EPSPS, mutations of which were being studied by this group, are not clear. The kinetics are tersely stated in the article itself:

‘Further evidence for correct delivery of the bacterial EPSP synthase into chloroplasts comes from the observation that plants expressing the fusion 3 polypeptide display a distinctive glyphosate tolerance phenotype readily distinguishable from the one conferred by cytoplasmic compartmentation of the aroA encoded EPSP synthase. Transgenic tomato plants expressing fusion 3 were tested in a field trial in Yolo County, California, in summer 1988. Application of 1 kg/ha of glyphosate salt (Roundup) at the 2-3 true leaf stage, or at the 5 to 6 true leaf stage, caused no loss in tomato yield. ’

On the face of it, these kinetics are apparently favourable, but without more information about the EPSPS used it is impossible to say what its qualities were. Professor Lichtenstein objected to the lack of data. He objected to the foregoing paragraph as a summary of the characteristics of the EPSPS. In any event, I am not satisfied that there was any evidence that this paper formed part of the common general knowledge, or even that class of documents that the skilled man would inevitably unearth confronted with a particular problem. So my conclusion is that it is not legitimate to rely on the Comai paper as a further disclosure encouraging the skilled person to seek out not mutants but tolerant/degrading microorganisms with any hope that they might display a favourable combination of kinetic qualities.

As it happens the Moore team missed the PG2982 EPSPS having concentrated on the gene responsible for glyphosate metabolism. Seven years after the publication of Moore the gene had not been identified.

My finding is that the skilled addressee at the priority date would have been presented with at least three main lines of attack on EPSPS, but would not have had much optimism about any of them, given glyphosate’s mode of operation by competitive binding. These lines of attack are mutation of known EPSPSs, over-expression and identification of resistant/degrading microorganisms obtained in nature, in the hope that one or more of them would show favourable kinetic characteristics. Moore is one publication. It does not describe the kinetic qualities of the EPSPS of PG2982, concentrating upon the enzyme responsible for degradation of glyphosate and failing to draw any inference as to the possible qualities of the EPSPS expressed in the microorganism. On the whole, and assuming in Cargill’s favour that PG2982 was available from the Moore group, I do not think that this publication renders the investigation of the EPSPS followed by identification of the gene, its cloning and expression obvious.

Fischer is an interesting paper. Its title is ‘Glyphosate sensitivity of [EPSPS] from Bacillus subtilis depends upon state of activation induced by monovalent cations’. The opening section of the paper includes this passage:

‘The recent intrigue about the genetic engineering of [glyphosate] resistance into plants has led to formulation of objectives for genetic alteration of EPSP synthases or identification of genes which code for [glyphosate] degradation. Another approach has been the search for EPSP synthase enzymes in nature that exhibit natural tolerance to inhibition by [glyphosate]. In this context, we noticed the apparent lack of sensitivity to inhibition by [glyphosate] displayed by EPSP synthase from Bacillus subtilis. The data reported here demonstrate that when EPSP synthase is characterized under what may otherwise be standard conditions of assay, the true sensitivity to inhibition by glyphosate may be masked. In particular, the cation requirements of EPSP synthases may be directly linked to sensitivity to [glyphosate]-promoted inhibition.’

At the end of the paper the importance of the results obtained to EPSPS assays is described:

‘If the cations contributed by the assay components are less than required for maximal activation, EPSP synthases of the B subtilis type would falsely appear to be relatively resistant to inhibition by [glyphosate] in vitro. Before assigning relative sensitivities of EPSP synthase enzymes to [glyphosate]-mediated inhibition, an evaluation of possible modulation of enzyme activity by monovalent cations seems essential. We also recommend the utilisation of Hepes/tetramethylammonium buffer plus the preparation of all substrates and [glyphosate] as tetramethylammonium salts if possible.’

Fischer is a warning to the skilled person: if you are not careful with the assay conditions, it may seem as though the EPSPS on which you are experimenting is glyphosate tolerant, when it is in fact sensitive. Under Fischer’s conditions, B subtilis EPSPS is glyphosate sensitive. The best that the evidence came to is the following, in re-examination of Professor Leaver (transcript 973):

In one sense, this is only the beginning. Assuming that ‘possibly acceptable’ gives the skilled person the impulse needed (I do not think it does) the sequence is not disclosed. Professor Leaver thought that Henner would appear on a literature search, if the term ‘aro’ were used: Henner does not refer to EPSPS. The evidence on the search was this (transcript 819-20):

This is the day-to-day labour of research. It is in my view to a high degree unfair to inventors to take the results of the work of this kind as obvious without considerably more. There is little point in confecting possible variant factual scenarios to demonstrate the point. While there is no doubt that what comes out by searching may well properly be described as obvious to the skilled man, what Professor Leaver describes in this passage is not the obtaining of information which should necessarily be taken to be obvious. The term ‘research project’ is useful, but conceals the fact that in any research this sort of dull, routine work has to be done. If in order to lay one’s hand on the B subtilis gene for EPSPS after reading Fischer these are the steps which have to be gone through in order to locate Henner and realize that the gene described by Henner is what you want, I do not think that this is to be viewed as an obvious step to take in the light of Fischer.

In my judgment, the allegation of obviousness on the basis of Fischer fails, not only because the exclusion of the Syngenta evidence means that it cannot be shown that B subtilis EPSPS is a Class II enzyme, but because it would not be obvious to take B subtilis further in the light of the Fischer disclosure.

‘Classical’ insufficiency of description is complained of in what I consider to be a central aspect of the patent, the identification of and definition of Class II EPSP synthases.

It will be recalled that claim 1 explicitly requires the enzyme to be ‘capable of reacting with antibodies raised against a Class II EPSPS enzyme selected from the group consisting of the enzymes of SEQ ID NO:3 and SEQ ID NO:5’, that is, CP4 and LBAA respectively. The first fundamental complaint is that this test is open-ended and that the skilled person cannot begin to be satisfied of a negative result without the expenditure of a wholly disproportionate scientific effort. In the ensuing discussion, it should be remembered that I am concerned not with the task of showing that something infringes, but with the task of showing affirmatively that it does not infringe.

Put shortly, the problems are said to be as follows.

The answer to at least (i) to (iv) comes, surprisingly, from the Syngenta experiments, the results of which I have excluded on grounds of fairness. Professor Lichtenstein, who strongly criticised the manner in which the experiments were conducted, was pressed to say whether he was also contending that the tests were incomplete (transcript 1032):

If the specification is insufficient, therefore, it must be because there are insufficient directions properly to design the experiments in detail: to use the correct controls, use crude test EPSPS to exclude non-specific cross-reactions etc. But I did not understand Professor Lichtenstein to suggest that the experimenters had been at sea, merely that they had not done what they should have done.

The Syngenta experiments had both a legal and a commercial purpose. If the experiments demonstrated that B subtilis EPSPS fell within the claim, then Syngenta considered that they would be able to show the patent was invalid. If, on the other hand, they failed to show that it fell within the claim, Syngenta considered that the enzyme would have been of interest: as Dr Hawkes said, a win/win situation. What everybody seems to have accepted is that the Syngenta experiments would, to this extent, have provided an answer. They were extensive and took much time to perform. But there was no question of Syngenta being unable to design them. I conclude that there is no insufficiency so far as (i) to (iv) are concerned.

There is a fundamental difficulty with (v), caused again by the lack of any experimental evidence from Cargill. There is an obvious potential weakness in the [0009] test which I consider inevitably part of the definition of a Class II enzyme: the range of Class I enzymes is wide and there is no direction as to which one to use in an assay. Nevertheless if those mentioned (Petunia, S typhimurium, E coli, tomato etc) raised antibodies some of which reacted with CP4 and some of which did not, it would be helpful if that were demonstrated, particularly given the specification’s insistence of the ‘remarkable degree of homology’ between them. Here a workable test is foreshadowed in [0034]. I think it would be most unsafe to assume that different results would be obtained with different enzymes as antigens as Cargill suggest.

Assuming, therefore, that the skilled person has performed the Syngenta sequence of experiments in an attempt to demonstrate non-infringement, and shown that there is no cross-reaction to antibodies raised to LBAA and CP4, is this result inconclusive? If it is, the specification is insufficient—see Kirin-Amgen v TKT [2005] RPC 9, [121]-[131], where there was no standard against which to test. I conclude that the evidence fails to establish that this is the case. The limit of the criteria of the claim to CP4 and LBAA, together with a failure to demonstrate that different results will be achieved with different anti-Class I EPSPS means that the claims are not open-ended and the tests, although onerous, can be performed.

In [0008], the patent says that ‘these kinetic constants [sc. K_m and K_i] are determined under the assay conditions specified hereinafter.’ The only reference to such conditions are in the part of the specification concerned with the purification of EPSPS isolated from CP4 in [0042] where it is said:

‘EPSPS enzyme assays were performed using either the phosphate release or radioactive HPLC method, as previously described in Padgette et al. 1987, using 1mM [PEP] and 2mM shikimate-3-phosphate (S3P) substrate concentrations. For radioactive HPLC assays, ¹⁴C-PEP (Amersham) was utilized.’

The pleaded case is that there is no direction as to which of these two tests should be used and that they give different results, and no directions as to cation identity or concentration, which are important to EPSPS activity (remember Fischer).

So far as the test to be used is concerned, there is really no difficulty. The cited paper expressly says that radioactive HPLC is to be used for kinetic determinations, and this Professor Leaver accepted, both in his report and in cross-examination.

The other point is more complicated. It is clear that the activity of EPSPS is capable of being affected by cation concentration, and that the directions in the Padgette et al 1987 paper permit pH to be adjusted by the addition of either KOH or NaOH. Cargill say the choice matters, again without any experimentation. Monsanto say that at the concentrations with which Padgette is concerned, it will not matter. There was a large amount of cross-examination, and an investigation of the Monsanto notebooks of one Diane Re, in which the manner in which Ms Re had performed the assays was gone over in some detail.

Padgette et al 1987 gives directions that the pH of the buffer, Hepes, (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) is adjusted to 7.0, but does not say how. If this point was to be made good, clear evidence of a relevant difference was required. Instead, Professor Leaver and Professor Bartlett gave rival interpretations of what Ms Re had done, Professor Leaver contending that she had problems with potassium concentration causing an increase in K_m when [K⁺] was increased, Professor Bartlett contending that this was not so: Funke, an article concerned with phosphate release assays, acknowledged to be less precise than HPLC, and Fischer, which has no data relating to the effect of sodium or potassium ion concentration on K_m, were relied on against this view.

While accepting the possibility of an effect, Professor Bartlett was unwilling to accept that it was ‘appreciable’, as there were no results and no figures, and pointed out that a good way of checking would be to compare a result with the patent: if the results were not comparable, the other cation should be used. I was not satisfied that this was a wrong approach.

The case put is two-fold: the claims cover sequences which are not glyphosate-tolerant to any useful extent, and there are insufficient criteria disclosed to enable the skilled person to identify any useful enzyme over and above the two (CP4 and LBAA) expressly disclosed. The basis of the objection is, of course, the speech of Lord Hoffmann in Biogen v Medeva [1997] RPC 1.

So far as the glyphosate tolerance of the claimed materials is concerned, it should be noted that Professor Leaver in his evidence was inclined to equate the requirements of the claim to commercially successful enzymes. If the proper view to take of the specification is that it provides a class of enzymes (subject to the next point) that are potentially usefully commercially but require further investigation, the position is different. I do not find in the specification any indication that all the enzymes having the qualities called for by claim 1 will necessarily be commercially useful, but merely that they will possess the degree of tolerance indicated by the kinetic requirements. In particular, the K_m requirement is wide, and the enzyme will have to be tested in individual plants so as to assess its performance when Roundup is administered.

I am far more concerned about the distinction between Class I and Class II drawn in the patent and reflected in the claim. Only two Class II enzymes appear to have been characterised at the date of the document, although at least two others are said to have been found, using a CP4 coding sequence probe: see [0068]. Throughout this case, I have been concerned at the lack of an exclusive criterion for the Class, the claim specifying an inclusive criterion. The draftsman has obviously and understandably eschewed the temptation to define his class in terms of its members’ ability to hybridise with the CP4 encoding sequence. His claim has five criteria for the protein which is encoded by the sequence: (1) it has EPSPS activity; (2) it reacts with antibodies raised to CP4; (3) (as I construe it) it does not react with antibodies raised to Class I; (4) 1µM  K_m  150µM; and (5) 3  K_i/K_m  500. If I can express the matter crudely, the patent spends much time ([0033]-[0034]) in pointing to structural homology between the various EPSPSs so far known on the one hand, and between CP4 and LBAA on the other. At the same time, it identifies a very ‘low degree of relatedness’ between CP4/LBAA on the one hand and the others. It also observes that sequences that are highly conserved in the putative Class I enzymes are not highly conserved between them and CP4/LBAA. See generally [0069]-[0072]. In this way, some structural colour is lent to the idea that there may be two structural families of enzymes, and this perception appears to be that which in part dictates the form of the claim of the patent.

I think that if the case put is that ‘the said claims are not limited to any principle which would enable the skilled person to identify glyphosate resistant EPSPS enzymes suitable for conferring glyphosate tolerance on plants and the specification discloses no such principle’, it is necessary to demonstrate, one way or another, that the defined class is bogus, its members being united, if at all, by the fact that they work and otherwise being properly described as a collection of single instances and the specification being silent as to how to find another. Here, the existence of a whole new class of enzymes is deduced on the basis of two exemplars only, which are admittedly similar. I can see that it might be difficult to demonstrate that if a larger collection of effective enzymes were examined it might appear that the supposed characteristics uniting them (general kinetics apart) were imaginary, but that seems to me to be the requirement. (By ‘general kinetics’ I mean the obvious desideratum that the binding rate to glyphosate must not be such that the growth of the plant is sufficiently impaired, or ‘it works in such-and-such a crop’.) I think this objection fails for this reason. Cargill have not demonstrated that the class is not a class.

I have set out the amended claims above in [25]. The first question is whether the specification discloses the Round Up Ready sequence, conveniently called CP4R, which differs from CP4 in the single S2L (serine to leucine at position 2) mutation I have described. Since the claims are all in one way or another intended to be limited to this sequence, no reference to it may be introduced if it is not already disclosed in the application for the patent.

This last point is important. In Triumph v Aeroquip-Vickers [2007] EWHC 1367 (Pat) [36]-[41] I gave my reasons for saying that the essential comparison was with the application as filed, and not with the specification prior to amendment, and I shall not repeat those reasons here. The nature of the comparison is described in Bonzel v Intervention (No 3) [1991] RPC 553 at 574 as follows:

‘The task of the Court is threefold:

(1)

to ascertain through the eyes of the skilled addressee what is disclosed, both explicitly and implicitly in the application.

(2)

To do the same in respect of the patent as granted.

(3)

To compare the two disclosures and decide whether any subject matter relevant to the invention has been added whether by deletion or addition. The comparison is strict in the sense that subject matter will be added unless such matter is clearly and unambiguously disclosed in the application either explicitly or implicitly.’

Finding the relevant disclosure is less straightforward than one might like. Starting with [0077], the fusion of the sequence for the chloroplast transit sequence from Arabidopsis thaliana (Footnote: ⁶) and CP4 is described.

‘The N-terminus of the CP4 EPSPS gene was modified to place a SphI site that spans the Met codon. The second codon was converted to one for leucine in this step also. This change had no apparent effect on the in vivo activity of the CP4 EPSPS in E coli as judged by rate of complementation of the aroA allele.’

A CTP sequence modified for SphI called CTP4 was then fused to the SphI-mutated CP4 sequence ([0079]). It is then incorporated into a plasmid vector called pMon13640, a map of which (fig 15) shows that this is, indeed, the vector with CTP4 and so one with mutated CP4, rather than CP4 as shown in the figure. Example 3 describes the transformation of plants with pMON13640 and the plant 13640/40-3 (Table X) is, as I understand it, the ancestor of all Round Up Ready soybean plants. [0097] says that the examples ‘are provided to better elucidate the practice of the present invention’ (sic).

What is disclosed is the transformation of a plant using recombinant CP4R, fused to a chloroplast transit protein sequence, and also the CP4R sequence prior to fusion.

There is no explicit disclosure that this sequence has any of the properties called for by the claim. The evidence is that it is overwhelmingly likely that it would satisfy both the kinetic and the immunological requirements. So is it necessarily implicitly disclosed as having those qualities? Because of the explicit statement in [0097], I think it is. This statement is necessarily to be interpreted as relating to the sequences with which the plant in Example 3 is transformed, which means CTP4 and CP4R.

The primary grounds upon which Cargill base themselves are that Monsanto has known of the alleged invalidities for too long. I agree that if any of these grounds had been made out, the question of the exercise of discretion so late in the patent’s life would give rise to difficult questions in the light of SKF v Evans Medical [1989] FSR 561, and the well-known obligation on patentees to amend promptly when substantive objections to validity are drawn to their attention if they do not wish to be embarrassed when seeking to avoid the consequences of a success in one of those objections later on.

As matters stand, however, the application becomes a voluntary restriction of the claims to the commercially successful embodiment of the invention, the sequence for Round Up Ready, disclosed as it is an example of this invention, although not separately characterised in any way. I cannot see any objection to such an amendment, and I grant leave for it to be made accordingly.

No claim of the patent is infringed by the importation of soybean meal on board MV Podhale. The patent is valid, and the application for permission to amend succeeds.