T 0162/86 (Plasmid p SG 2) of 7.7.1987

Summary of Facts and Submissions

I. European patent application No. 82 103 443.6, filed on 23 April 1982 claiming priority of 30 April 1981, was refused by adecision of the Examining Division dated 3 February 1986. This decision was based on two sets of claims: Claims 1 to 5, received on 28 January 1985, for all designated Contracting States except Austria, and Claims 1 to 4, also received on 28 January 1985, for Austria. (a) Claim 1 for the Contracting States Belgium, France, Germany, Italy, Liechtenstein, Luxembourg, the Netherlands, Sweden, Switzerland and the United Kingdom read as follows: "1. Plasmid pSG2, obtainable from Streptomyces ghanaensis ATCC 14 672, characterised by a molecular weight of 9.2 megadaltons, a contour length of 4.58 µm and a molecular length of about 13.8 kb." (b) Claim 1 for Austria read as follows:

"1. Process for obtaining plasmid pSG2, characterised in that a culture of Streptomyces ghanaensis ATCC 14 672 is lysed and the plasmid isolated."

In both sets of claims the other claims related to the production of plasmid pSG2 or its use to construct vector plasmids.

II. The application was refused on the grounds that while the subject-matters of these claims were novel they were not based on inventive step. The problem which the disputed application set out to solve was to obtain sufficient quantities of a plasmid suitable as a vector for the genetic improvement of streptomyces strains. However, the proposed solution, i.e. the claimed plasmid pSG2, did not produce any noticeable (genetic) improvement when compared with the plasmid pUC6 known from DE-A-3 005 226 (i). Nor did the use aimed at here represent an inventive step, since such a use was inherent in the nature of plasmids which, according to their scientific definition, are used to make vector (hybrid) plasmids, as was also known from document (i). Furthermore, it was indisputably known from the state of the art that plasmids are to be found in species of streptomyces. It therefore automatically followed that the person skilled in the art could expect plasmids to be present in species of streptomyces such as the known Streptomyces ghanaensis. The Examining Division, moreover, did not know of any species of streptomyces which did not contain plasmid.

Claim 1 would not be allowable under Article 56 EPC in any event since in the present case the mere process of isolating a plasmid from Streptomyces ghanaensis would not warrant acceptance that the plasmid itself involved an inventive step.

III. The appellants (applicants) filed an appeal (received on 27 March 1986) against the aforementioned decision and paid the required fee. Their statement of grounds was received on 2 May 1986.

IV. Oral proceedings took place on 7 July 1987. The case made by the appellants in the course of written and oral procedure was essentially as follows: Document (i) said nothing about the host range of pUC6. In this respect the claimed plasmid brought about a "genetic improvement" by the mere fact that, as experience had shown, the hybrid vectors produced from it could be expected to be particularly stable in Streptomyces ghanaensis host cells on account of the endogenous component. In contrast to document (i), the problem was really to make available a plasmid, suitable as a (hybrid) vector, which was not immediately eliminated by the streptomyces strain used as the host cell. This problem was solved by the claimed endogenous plasmid pSG2 isolated from Streptomyces ghanaensis (ATCC 14 672). Since the state of the art did not give the slightest indication that plasmids were to be found in Streptomyces ghanaensis in the first place, let alone in one particular strain, the Examining Division's finding that it was not aware of any streptomycetes that did not have plasmids was, in view of the Declaration received in the examination proceedings in question on 28 January 1985, absolutely untenable. In reality, the occurrence of plasmids in streptomycetes was relatively rare, as the expert report filed on 2 May 1986 confirmed. In the course of the examination proceedings the appellants had also pointed out that the streptomyces strains specified in DE-C-1 113 791 (II) alongside Streptomyces ghanaensis did not, as far as they were aware, contain any plasmids (see page 2, second paragraph, of letter received on 28 August 1985).

V. The appellants requested that the decision refusing the application be set aside and a patent granted on the basis of the documents submitted at the oral proceedings. (a) Claim 1 (for all designated Contracting States except Austria) now reads as follows:

"1. Plasmid pSG2, characterised in that it is obtainable from a culture of Streptomyces ghanaensis ATCC 14 672, has a contour length of 4.58 µm and a molecular length of about 13.8 kilobases (kb) and in that it is not fragmented by the restriction endonucleases Eco RI, Bam HI, Sal I, Hpa I and Hin dII, but is cleaved by the restriction endonuclease Hin dIII into a fragment with a length of about 14 kb, by Cla I into two fragments with lengths of 10.15 and 3.65 kb, by Pst I into two fragments with lengths of 10.85 and 3.0 kb, by Bgl II into two fragments with lengths of 11.25 and 2.6 kb, and by Bcl I into three fragments with lengths of 11.6, 1.25 and 1.0 kb." (b) Claim 1 for Austria now differs from this only in that the product claim has been changed into a process for obtaining plasmid pSG2. Otherwise it is identical. All the other claims in both sets are identical to the claims with the same numbers on which the decision to refuse the application was based.

Reasons for the Decision

1. The appeal complies with Articles 106 to 108 and Rule 64 EPC and is therefore admissible.

2. For reasons of clarity, it seems imperative to include alongside the characteristics given in the rejected Claim 1 other characteristics within the limits of the original disclosure, in order to ensure unequivocal identification of the claimed plasmid. Besides details of origin (strain), contour length, molecular weight and/or molecular length, the fragmentation behaviour on digestion by various enzymes must be considered essential in this connection, particularly since determining the molecular weight and/or molecular length is fraught with inaccuracy, and also since in the presence of several plasmids which may be derivatives of one and the same plasmid and which possibly had not even been recognised originally, it must at all times be clear which plasmid is protected by the claim. However, it is not absolutely necessary for this purpose to include in the claim all the fragmentation results given in the original description, especially not if such details are lacking in accuracy. The present Claim 1 fulfils these criteria and therefore satisfies the conditions of Article 84 EPC.

3. Neither of the two versions of the claim can be objected to under Article 123(2) EPC. The present Claim 1 (for all designated Contracting States except Austria) was produced by combining the original Claims 1 and 2 and deleting the molecular weight of the plasmid expressed in megadaltons. This deletion merely involves the omission of an item of information which can in any case be derived by a person skilled in the art from the molecular length in kilobases given in the claim. The same goes for the redrafted Claim 1 for Austria as far as the characteristics of the plasmid are concerned. The other characteristics of this claim, which relate to the process, are supported in the original Claim 3.

4. The subject-matter of the disputed application concerns plasmid pSG2 and a method for isolating it from a culture of Streptomyces ghanaensis ATCC 14 672.

5. The state of the art coming closest is document (i), which describes the isolation of plasmid pUC6 from a culture of Streptomyces espinosus (NRRL 11 439). This plasmid, which is likewise obtained from a streptomyces strain, has a molecular length of approximately 9.2 kilobases and a restriction map as indicated in the drawing. It is suitable as a cloning vehicle and can be used to create recombinant plasmids which, through transformation, can be inserted into host bacteria (see page 6, first and second paragraphs; page 18, last paragraph and drawing).

The advantage of plasmid pUC6 lies in its ability to act as a plasmid vector in industrially important micro-organisms, for example streptomycetes. Thus by cloning genetic information from streptomycetes into pUC6 it is possible to increase the output of industrially valuable products such as antibiotics from these organisms. Since this is a streptomyces plasmid, it is ideally suited as a vector for the streptomyces genus but can also be used as a vector in other micro-organisms (e.g. bacillus, arthrobacter) (see page 20, first paragraph and page 21, lines 9 to 11).

6. With reference to document (i), however, the appellants have stressed that, as far as they know, neither the strain mentioned nor the genus S. espinosus has attained any technical importance. There is nothing to indicate that this is not the case, and the Board therefore has no reason to doubt this statement. However, when vectors based on the prior art plasmid pUC6 are used in species other than S. espinosus which are of industrial importance in the production of antibiotics for example, elimination problems are a source of failure on acount of the necessarily exogenous nature of the original plasmid in these cases.

7. In contrast to document (i), the problem is therefore to make available a plasmid that, when used as a hybrid vector, is not immediately eliminated again by the streptomyces strain used as the host cell, i.e. is stable in the host. With regard to the solution proposed (plasmid pSG2), the appellants had already pointed out in the examination proceedings that, compared with plasmid pUC6 known from (i), plasmid pSG2 at least had the advantage that, as an endogenous Streptomyces ghanaensis plasmid, it is stable in this host even in the form of hybrid plasmids, thanks to its endogenous component (see submission received on 28 August 1985, page 5, first paragraph). It is this which forms the basis of the "genetic improvement" of the claimed plasmid asserted in the appeal proceedings. The Board's view is that this only serves to clarify the information given in the original description that plasmid pSG2 was inter alia a suitable starting plasmid for the application of genetic engineering methods to the Streptomyces ghanaensis strain itself (see page 2, lines 7 to 22 of the description). On the basis of established case law, the only yardstick for objectively defining the problem is what is actually achieved vis-à-vis the state of the closest art. Although the general problem to be solved according to the original description consisted in obtaining sufficient quantities of a plasmid suitable as a vector for the genetic improvement of streptomycetes, there is nevertheless nothing to prevent a more exact definition of the original problem from being given, even in the course of the appeal, provided it is supported by the original disclosure (cf. Decision T 184/82, "Poly(p-methyl- styrene) articles/MOBIL", OJ EPO 1984, 263, 264, in particular Reasons for the Decision 3 to 5). Such a restatement of the problem is admissible in the present case since, as explained further above, the advantage invoked by the appellants is already implied in the original problem.

8. According to product Claim 1, the proposed solution to the existing problem is plasmid pSG2, which comes from a culture of Streptomyces ghanaensis ATCC 14 672 and is additionally characterised by a series of further features stated in the claim, such as contour length, molecular length and fragmentation behaviour (cleavage pattern) when digested by various enzymes. What has to be examined is whether the problem stated is solved by the claimed subject-matter, the problem being more specifically the stability of the plasmid when used as a vector in the manufacture of antibiotics.

Document (i) does not contain any information about the technical significance of the species S. espinosus and the skilled person thus has to ask himself what use such a micro-organism could have as a host. He knows only from other streptomycetes that they are suitable for the manufacture of antibiotics. It is known, for instance, from document (II) that a strain of S. bambergiensis or other equivalent streptomycetes such as S. ghanaensis, S. ederensis or S. geysiriensis can be used to manufacture the antibiotic Moenomycin (see patent claim).

In document (II) obviously no use is made of genetic engineering methods. If, however, the skilled person wished, as in the present case, to use in conventional manufacturing processes such new techniques as are described in document (i), previously all that was available to him was the exogenous plasmid pUC6 described in that document obtained from S. espinosus, for which the question of compatibility with potential host organisms inevitably arises.

The improvement in the production of antibiotics which experts have endeavoured to achieve by this route inter alia can, however, only be achieved satisfactorily if the vectors needed for this can be permanently integrated into the micro-organism. By their very nature, vectors based on exogenous plasmids do not offer the best prerequisites for this. Furthermore, there would be little point in using a host organism in which the vector did have the necessary stability but could only produce a product for which there was no demand whatever or, in an extreme case, could only manufacture a product which was completely unusable (e.g. for reasons of toxicity). However, since a starting plasmid suitable for such purposes has now been obtained from the technically useful strain S. ghanaensis and an endogenous starting plasmid is thus available for a host which is suitable for manufacturing antibiotics, the Board is of the opinion that the problem described has been convincingly solved, because the person skilled in the art can expect endogenous plasmids to lead to highly stable hybrid plasmids.

9. The Examining Division affirmed that plasmid pSG2 was novel. Likewise the Board, having examined the documents cited in the course of the proceedings, has not discovered anything prejudicial to the novelty of the claimed plasmid pSG2. The subject-matter of Claim 1 (for all the designated States except Austria) is therefore novel.

10. The claimed solution must therefore be examined to see whether it is also based on inventive step.

10.1 The plasmid pUC6 known from document (i) is obtained from Streptomyces espinosus (NRRL 11 439) which is particularly suitable as a cloning vehicle for the streptomyces genus and thus opens up the possibility of increasing the yield of industrially useful products (e.g. antibiotics) from these organisms. No details of the host range of such vectors are given in document (i), but the document is obviously based on the assumption of wide use in this connection within the streptomyces genus and even to some extent outside this genus (in bacillus, etc.). However, this means that such a vector for the selected host usually consists of an exogenous plasmid and it is therefore to be expected that the vector will not always be stable in the host cell.

10.2 It is accepted that the danger of elimination can be reduced if for construction of the vector to be inserted into the host cell an endogenous plasmid is available which meets the requirements necessary for such purposes (molecular weight, copy number, etc.). If all bacteria, or at least all streptomycetes, were to contain aplasmid, the skilled person would then normally have no difficulty isolating the endogenous plasmid for any host micro- organism of the streptomyces genus, and all that would remain to be done would be to investigate whether the plasmid structure was suitable for use in genetic engineering. However, plasmids are by no means ubiquitous, as is perfectly clear from the appellants' Declaration, submitted during the examination procedure, which specifically states that, in a citation published before the priority date, of 34 (streptomyces) strains tested only 4 contained plasmids, and in another citation (likewise published before the priority date) of 32 strains tested only 7 contained supercoiled DNA (which can be an indication of plasmids) (loc.cit. 4 (c) and (d)). These statements of the appellants, which in themselves are convincing, are further corroborated by the supporting document submitted on 2 May 1986 (abridged version of a lecture given in 1985 - (III)) from which it emerges that of 127 strains of streptomyces from 4 species only 9 contained a plasmid. The appellants thus provided convincing proof of the extremely heterogeneous character of this genus as far as the occurrence of plasmids is concerned. In contrast, the unsupported objection of the Examining Division that, to date, it did not know of any species of streptomyces which did not contain plasmid is not convincing and appears to be uninformed. In an earlier unpublished decision, another Board did not accept an unsubstantiated objection as to novelty which was based exclusively on personal knowledge (cf. T 21/83, in particular point 4 of the Reasons for the Decision). Although in this instance the question of novelty is not at issue, the Board nevertheless - in the light of the aforementioned Declaration alone - has no reason to take a different view in assessing inventive step. As matters stand, the Board must therefore assume that the occurrence of plasmids in streptomycetes is sporadic. It accordingly has no reason to doubt that the other streptomyces strains named in document (II) along with Streptomyces ghanaensis are all devoid of plasmids. Assuming the skilled person had tried to apply the new genetic engineering methods to the conventional process described in document (II), he would in all probability have chosen the strain of S. bambergiensis mentioned first in the document, the only one, incidentally, for which an example is given in the description. Knowing full well that from the outset there would be very little chance of finding a strain containing plasmid and that not every plasmid found is suitable for genetic engineering applications (see below), he would only have been able to establish through such an experiment the absence of plasmids in that micro-organism. This not unexpected finding would certainly not have encouraged him to follow this course of action further, however. From this it does conclusively follow though that the cited state of the art could not give any grounds for believing that Streptomyces ghanaensis ATCC 14 672 contained a plasmid, let alone that this plasmid is constituted in such a way that it is suitable for genetic engineering applications, e.g. for the construction of vectors.

10.3 In order to solve the problem, a strain within the genus streptomyces would have to be found which contained plasmid, an undertaking which because of the sporadic occurrence of plasmids in streptomycetes could not be approached methodically. In addition, the plasmid sought would also have to be indeed suitable for the intended purpose. According to the appellants' statements, it may be assumed that plasmids for genetic engineering purposes should have a molecular weight of less than about 17 Md (equivalent to 25 kb). It is known that not all plasmids conform to this requirement. The claimed plasmid pSG2 has a molecular weight of 9.2 Md (about 13.8 kb) and thus lies slightly above that of the prior art plasmid pUC6 (6.0 Md). The Board could find nothing to suggest that flawless operation would no longer be possible as a result, or that other major disadvantages would arise. Furthermore, finding suitable plasmids is not a simple routine matter, as is quite evident from the aforementioned document (III). If of 127 strains from 4 species only 9 contained a plasmid, and of these only 3 had a molecular weight of less than 17 Md (less than 25 kb), this in fact means that of all the strains examined only agood 2% contained a plasmid of suitable molecular size at all. Moreover, there are apparently also species in which no plasmids could be detected, despite intensive searching (e.g. in the 31 strains of S. albus tested). A plasmid for genetic engineering purposes must not only have a suitable molecule size but must in addition be manipulatable, analysable and also identifiable or distinguishable from other plasmids. The plasmid's restriction behaviour on digestion by various enzymes is of importance in this connection. As is already clear from the original description (see Table, page 3), however, plasmid pSG2 is, if anything, superior to plasmid pUC6 in this respect because of the larger number of double cleavage sites (3:1), while each has only one unique cleavage site. In this connection the appellants have expressly pointed out both in the examination proceedings (see page 5, first paragraph, of letter received on 28 August 1985) and on appeal (page 3, paragraph 2 of the statement of grounds) that after the filing date further unique cleavage sites were found, so that at the moment four such cleavage sites are known, namely for Hin dIII, Eco RV, Pvu II and Nhe I. Since unique cleavage sites occur only rarely, this has to be seen as a further sign of the eminent suitability of plasmid pSG2 for genetic engineering purposes. In document (i) only one unique cleavage site (Bgl II) is indicated for the plasmid pUC6. It is not known whether further unique cleavage sites were subsequently found. The fact that a skilled person would have been aware from the technical literature of a number of other streptomyces strains containing plasmids could not have helped him in the search for a plasmid suitable for genetic engineering purposes, since he would not thereby have obtained any concrete information as to where within the genus he would have to look in order to be successful in the present case.

10.4 The solution to the problem stated therefore comes directly from the non-obvious choice made within the genus streptomyces. The slightly higher molecular weight of the claimed plasmid compared with plasmid pUC6 could not be shown to have any really serious disadvantage, although plasmid pSG2 consequently comes closer than the prior art plasmid to the critical limit of 17 Md. Nor can the difference in the copy number, which at 20 to 40 for pUC6 is roughly twice as high as for the claimed plasmid (10 to 20), be definitively pronounced a disadvantage or an advantage. While a higher copy number is an advantage for the isolation of the plasmid, it can nevertheless have fatal consequences for the host cell under certain circumstances (see Declaration, point 2). These differences can therefore be disregarded. They are in any case not relevant to the assessment of inventive step since they are in no way pertinent either to the problem or its solution. The subject-matter of Claim 1 (for all designated States except Austria), i.e. the claimed plasmid pSG2, is therefore based on inventive step.

11.1 The associated process in accordance with Claim 2 and the subsidiary Claims 3 and 4 which are dependent thereon is what is known as an analogy process, which was used for the first time in the application at issue to isolate a plasmid from a culture of Streptomyces ghanaensis ATCC 14 672. The subject-matter of Claim 2 is therefore novel. The same goes for Claims 3 and 4 which are dependent on it. Likewise, the use of plasmid pSG2 to construct vector plasmids has obviously not been described before and is therefore novel.

11.2 As with other areas of chemistry, inventive step also has to be acknowledged here, since it is evident from the inventive product, namely from plasmid pSG2 which is suitable for constructing a vector which is stable in the host organism.

12. The above findings also apply to the process Claims 1 to 3 and the use Claim 4 for Austria.

13. The subject-matter of both sets of claims is therefore novel and is based on inventive step. Both sets of claims are therefore allowable (Article 52(1) EPC).

ORDER

For these reasons, it is decided that:

1. The contested decision is set aside.

2. The case is referred back to the department of first instance with the order that a European patent be granted on the basis of the following documents submitted at the oral proceedings:

-Claims 1 to 5 for the Contracting States Belgium, France, Germany, Italy, Liechtenstein, Luxembourg, the Netherlands, Sweden, Switzerland and the United Kingdom;

-Claims 1 to 4 for Austria;

-Description comprising pages 1 to 8.