T 1336/04 (Cellulase/NOVOZYME) 09-03-2006

Summary of Facts and Submissions

I. European patent No. 0 531 315 with the title "An enzyme capable of degrading cellulose or hemicellulose" was granted on the basis of European patent application No. 91 908 986.2 with 18 claims for all designated Contracting States except Spain (ES) and with 10 claims for ES. The patent was opposed on the grounds of Articles 100(a), (b) and (c) EPC. On 29 January 2001, the opposition division revoked the patent because the granted claims were considered to contravene Article 123(2) EPC.

II. The patentee lodged an appeal against the interlocutory decision of the opposition division. In the course of appeal proceedings, the appellant maintained the claims as granted as main request and filed auxiliary requests 1 and 2 (AR1, AR2). The then competent board decided that the main request and AR1 did not fulfil the requirements of Article 123(2) EPC. These requirements were found to be fulfilled by AR2 and, pursuant to Article 111 EPC, the board remitted the case to the opposition division for further prosecution on the basis of this auxiliary request (cf. T 345/01 of 14 February 2003).

III. In its interlocutory decision of 20 August 2004, the opposition division decided that the main request, which corresponded to the AR2 underlying decision T 345/01 (supra) and auxiliary requests 1 to 3 filed on 30 January 2004 did not fulfil the requirements of Article 56 EPC. Furthermore, the main request and the auxiliary requests 1 and 2 were found not to fulfil the requirements of Article 83 EPC.

IV. An appeal was lodged by the patentee (appellant), who filed the statement of grounds of appeal on 21 December 2004 and requested that the decision under appeal be set aside and that the patent be maintained in amended form on the basis of the main request and the auxiliary requests 1 to 3 that had been considered by the opposition division in the decision under appeal. On 13 May 2005, the opponent (respondent) replied to the grounds of appeal.

V. Summons to oral proceedings were sent on 20 October 2005. In a communication annexed thereto (Article 11(1) of the Rules of Procedure of the Boards of Appeal, OJ EPO 2003, 89), the parties were informed of the board's preliminary opinion.

VI. On 9 February 2006, both appellant and respondent filed further observations in reply to the board's communication.

VII. Oral proceedings took place on 9 March 2006. During oral proceedings the appellant withdrew all auxiliary requests and filed new auxiliary requests 1 and 2.

VIII. Claim 1 of the main request for all designated Contracting States except ES read as follows:

"A cellulose- or hemicellulose-degrading enzyme which is derivable from a fungus other than Trichoderma or Phanerochaete, and which comprises a carbohydrate binding domain homologous to a terminal A region of Trichoderma reesei cellulases, wherein the carbohydrate binding domain comprises the following amino acid sequence

Trp Gly Gln Cys Gly Gly Gln Gly Trp Asn Gly Pro Thr Cys Cys Glu Ala Gly Thr Thr Cys Arg Gln Gln Asn Gln Trp Tyr Ser Gln Cys Leu;

Trp Gly Gln Cys Gly Gly Ile Gly Trp Asn Gly Pro Thr Thr Cys Val Ser Gly Ala Thr Cys Thr Lys Ile Asn Asp Trp Tyr His Gln Cys Leu;

Trp Gly Gln Cys Gly Gly Ile Gly Phe Asn Gly Pro Thr Cys Cys Gln Ser Gly Ser Thr Cys Val Lys Gln Asn Asp Trp Tyr Ser Gln Cys Leu;

Trp Gly Gln Cys Gly Gly Asn Gly Tyr Ser Gly Pro Thr Thr Cys Ala Glu Gly - Thr Cys Lys Lys Gln Asn Asp Trp Tyr Ser Gln Cys Thr Pro;

Trp Gly Gln Cys Gly Gly Gln Gly Trp Gln Gly Pro Thr Cys Cys Ser Gln Gly - Thr Cys Arg Ala Gln Asn Gln Trp Tyr Ser Gln Cys Leu Asn;

Trp Gly Gln Cys Gly Gly Gln Gly Tyr Ser Gly Cys Thr Asn Cys Glu Ala Gly Ser Thr Cys Arg Gln Gln Asn Ala Tyr Tyr Ser Gln Cys Ile;

Trp Gly Gln Cys Gly Gly Gln Gly Tyr Ser Gly Cys Arg Asn Cys Glu Ser Gly Ser Thr Cys Arg Ala Gln Asn Asp Trp Tyr Ser Gln Cys Leu;

Trp Ala Gln Cys Gly Gly Asn Gly Trp Ser Gly Cys Thr Thr Cys Val Ala Gly Ser Thr Cys Thr Lys Ile Asn Asp Trp Tyr His Gln Cys Leu;

Trp Gly Gln Cys Gly Gly Gln Asn Tyr Ser Gly Pro Thr Thr Cys Lys Ser Pro Phe Thr Cys Lys Lys Ile Asn Asp Phe Tyr Ser Gln Cys Gln; or

Trp Gly Gln Cys Gly Gly Asn Gly Trp Thr Gly Ala Thr Thr Cys Ala Ser Gly Leu Lys Cys Glu Lys Ile Asn Asp Trp Tyr Tyr Gln Cys Val

or a subsequence thereof capable of effecting binding of the enzyme to an insoluble cellulosic or hemicellulosic substrate."

Claims 2 to 6 were embodiments of claim 1, wherein claim 3 defined eight further amino acid sequences comprised in the linking B region that connected the carbohydrate binding domain (CBD) to the catalytically active domain of an enzyme of claims 1 or 2. Claims 7 to 10 related, respectively, to a DNA construct comprising a DNA sequence encoding an enzyme of any of claims 1 to 6, to an expression vector carrying said DNA construct, and to a cell transformed with this DNA construct or expression vector. Claim 11 was directed to a method of producing the enzyme of any of claims 1 to 6 using transformed cells. Claims 12 and 13 were concerned with an agent for degrading cellulose or hemicellulose comprising an enzyme of any of claims 1 to 6. Claims 14 and 15 were directed, respectively, to specific amino acid sequences comprised in the CBD (and subsequences thereof) and in the linking B region.

IX. Auxiliary request 1 (AR1) for all designated Contracting States except ES read as the main request except for the deletion of claims 4, 5 (with consequent renumbering) and of claims 14 and 15 and the limitation of claim 1 to the eighth amino acid sequence of claim 1 in the main request.

X. Auxiliary request 2 (AR2) for all designated Contracting States except ES read as AR1 except for the deletion of references to any subsequence in claim 1.

XI. The corresponding sets of claims were filed for ES.

XII. The following documents are mentioned in the present decision:

(3): J. Knowles et al., Trends in Biotechnology, 1987, Vol. 5, pages 255 to 261;

(4): P. Sims et al., Gene, 1988, Vol. 74, pages 411 to 422;

(11): WO-A-89/09259 (publication date: 5 October 1989);

(15): M. de O. Azevedo and A. Radford, Nucleic Acid Res., February 1990, Vol. 18(3), page 668;

(16): "Description of PASC assay", filed with patentee's letter dated 30 January 2004;

(17): A report from the patent proprietor filed with submissions dated 21 December 2004 and entitled: "In silico hybridization of three DNA sequences to the Humicola insolens Endoglucanase V cDNA sequence", pages 1 to 17;

(19): "Chambers Twentieth Century Dictionary", Ed. by A.M. Macdonald BA(Oxon), edition 1972, pages 20 and 412.

XIII. Appellant's arguments in writing and during oral proceedings insofar as relevant to the present decision may be summarised as follows:

Main request

Article 123(2) EPC

According to the established case law, the same binding effect applied to a subsequent appeal in respect of an earlier decision of a board of appeal as it applied to the department of first instance. Insofar as the facts were the same, the board was, thus, bound by the ratio decidendi of the earlier decision T 345/01 (supra). Since the main request did not differ from the AR2 found to fulfil the requirements of Article 123(2) EPC by the then competent board of appeal, the issue of conformity with this article was res judicata and may not be investigated anymore.

Article 54 EPC; Claim 1 relating to enzymes with a CBD comprising a subsequence capable of effecting binding of the enzyme to an insoluble cellulosic or hemicellulosic substrate.

Document (19) showed that the term "affect" meant "to act upon or to influence" in contrast to the term "effect" which meant "to produce; to accomplish; to bring about or to cause". Hence, the CBD subsequence referred to in the claim had to be capable of accomplishing the binding to cellulose. The respondent's argument that a subsequence capable of influencing (affecting) the binding was comprised within the claim implied that said subsequence could be as short as a single amino acid since individual amino acids were expected to contribute to the binding and thus, it was not technically sensible. The CBD subsequences identified by the respondent as being in documents (4) and (15) as well as in the CBD sequences of the claims had not been shown to effect the binding to cellulose and, therefore, could not be detrimental to novelty.

Article 56 EPC; Claim 1

The closest prior art document (3) gave the skilled person an incentive to look for new cellulases and genes encoding them. However, document (3) gave no guidance as to how to isolate them. The document described the presence of homologous terminal domains within four cellulases from Trichoderma reesei but there was no indication that similar domains were to be found in other cellulases. In fact, these homologous terminal domains from Trichoderma were described as being different from the domains found in two cellulases from Cellulomonas. Document (3) referred to the great variety in the structure of different cellulases and to their very dissimilar primary structure even when comparison was made among cellulases derived from a single organism. The comparison of carbohydrate binding proteins with known three dimensional structures showed that there was a surprising amount of diversity in their tertiary structures and, therefore, in their CBDs as well.

Since the activity of the cellulose-degrading enzymes was defined by their catalytic domains, which were of much larger size than the CBDs, the skilled person would have obviously used the full-length sequence of the known cellulase genes or oligonucleotide probes derived from these catalytic domains for screening cDNA libraries. There was no hint in the prior art inciting the skilled person to use an oligonucleotide probe derived only from the CBD.

Document (4) reported the cloning of cellobiohydrolase I (CBHI) gene from Phanerochaete chrysosporium using as a probe a genomic fragment from T. reesei which encompassed half of the coding region of the CBHI gene (Figure 1). However, there was no evidence on file showing that this probe comprised a sequence encoding the CBD. Although document (4) referred to the homology between the C-termini of the CBHI enzymes from P. chrysosporium and T. reesei, the comparison of their complete amino acid sequences showed other regions with homology, in particular in their catalytic domain (Figure 9). There was no suggestion that homologous terminal CBDs were to be found in all cellulases nor an indication that probes derived therefrom could be of any particular relevance. In fact, the presence of homologous CBDs in other fungi could not be directly predicted from document (4), since their presence in both the phylogenetically unrelated P. chrysosporium and T. reesei could be due to a recent horizontal genetic exchange between these fungi.

Document (15) disclosed the cloning of the CBHI gene from Humicola grisea var. thermoidea using as a probe essentially the entire coding region of the CBHI gene from P. chrysosporium. Although the putative amino acid sequence encoded by the CBHI gene was disclosed, there was no reference to the domains of the enzyme nor to the presence of any homology with other CBHI enzymes or domains thereof.

Thus, the selection of an oligonucleotide probe derived from the CBD could not be plainly inferred from this prior art and the cloning strategy of the patent in suit was not obvious to the skilled person. A contrario, there was evidence on file showing that a probe derived from the CBD was essential for obtaining the enzymes of the patent.

Document (17) (cited as expert opinion) showed that probes directly derived from the prior art, i.e. CBHI genes from T. reesei, H. grisea or P. chrysosporium, did not hybridize to a cDNA sequence encoding the endoglucanase V (EGV) from H. insolens which comprised the eighth amino acid sequence of claim 1. Since all sequences of claim 1 were closely related, the results obtained for EGV could be reasonably extrapolated to the other sequences of claim 1. None of these sequences could thus be isolated using the probes directly available from the prior art.

As for document (16), it showed that the EGV from H. insolens was unexpectedly more active (145%) than other known cellulases, such as the EGI from T. reesei. Since all sequences of claim 1 were closely related, the surprising results obtained for the EGV could be reasonably extrapolated to all other sequences. Although filed in 2004, this evidence could be taken into account in the assessment of inventive step because it supplemented the technical contribution disclosed in the patent. The present situation was thus different from the one underlying decision T 1329/04 of 28 June 2005, in which inventive step was denied because it was solely on the basis of post-published evidence that the then claimed subject-matter could be identified as a bona fide solution to the problem to be solved.

In summary, the patent in suit disclosed an inventive cloning strategy using a non-obvious probe and also provided advantageous cellulases that were themselves inventive. Claims concerned with specific CBDs or linker sequences were also inventive, since they were directly derived from this cloning strategy and they were only "part" of the complete (inventive) cellulases.

Auxiliary requests 1 and 2

Article 56 EPC; Claim 1

The subject-matter of claim 1 of both these requests was limited to a cellulose- or hemicellulose-degrading enzyme comprising a CBD comprising the eight amino acid sequence shown in claim 1 of the main request (with or without subsequences derived therefrom, respectively, AR1 or AR2). Evidence was on file showing that an endoglucanase comprising this specific CBD had an advantageous higher activity. It was plausible that this specific CBD also improved the performance of other cellulases, since lower cellulose-degrading activity was found with other CBDs. The unexpected effect disclosed in document (16) could reasonably be extrapolated to all cellulases that comprised this specific CBD. There was no evidence on file showing that such extrapolation was wrong or not reasonable. Thus, inventive step was to be acknowledged on the basis of this advantageous effect.

XIV. Respondent's arguments in writing and during oral proceedings insofar as relevant to the present decision may be summarised as follows:

Main request

Article 123(2) EPC

Article 100(c) EPC was a ground for opposition from the beginning of the proceedings. It was always within the discretion of the board to decide whether an argument was so relevant as to need to be considered. This was the case for the argument under Article 123(2) EPC which was presented in the course of this appeal.

Article 54 EPC; Claim 1 relating to enzymes with a CBD comprising a subsequence capable of effecting binding of the enzyme to an insoluble cellulosic or hemicellulosic substrate.

The binding of cellulose to a CBD was a quantitative effect. However, there was no indication in the claims nor in the description, of the degree of binding required for the CBD subsequences referred to in the claims to be considered as being able to bind to cellulose. In the absence of any indication, the only technically sensible interpretation of the term "capable of effecting binding" was that these CBD subsequences only had "to contribute" or "to play a role" in the binding to cellulose.

Document (15) disclosed the CBHI gene of H. grisea var. thermoidea encoding a cellobiohydrolase. The CBD at the C-terminus of the amino acid sequence comprised several subsequences which were also found in the CBD sequences of claim 1. Since these subsequences were conserved in all known CBD, it was reasonable to assume that they played an important role in effecting the binding to cellulose in the same manner as the CBDs disclosed in the patent in suit were assumed "to effect" the binding of the enzyme to cellulose on the basis of sequence homology with known CBD sequences. The teaching of document (15) was thus detrimental to the novelty of claim 1.

Article 56 EPC; Claim 1

The closest prior art document (3) referred to four cellulases from T. reesei having homologous terminal domains and the same basic architecture as other known cellulases (catalytic domain and CBD joined by a linking B region). Although there was no reference to the presence of these homologous domains in cellulases from other organisms, the attention of the reader was nevertheless drawn to a document (document (4) on file) disclosing that a gene had been isolated from P. chrysosporium which was similar but not identical to the T. reesei CBHI gene. In fact, the person skilled in the field of cellulose-degrading enzymes was well aware of documents (4) and (15) which disclosed cellulases with homologous terminal CBDs.

Starting from the closest prior art, the problem to be solved was the provision of alternative cellulases comprising a CBD homologous to the ones from T. reesei but different therefrom. Claim 1 purportedly provided cellulases comprising those homologous CBD sequences. The problem was not, however, credibly solved since the activity of the alleged cellulases was not shown in the patent and it could not be directly derived from the disclosed CBDs, which were not the catalytic domain of these enzymes.

Homologous terminal CBDs were known to be present in the cellulases disclosed in documents (3), (4) and (15). The selection of probes based on these homologous CBDs was almost a "one-way-street" situation associated with a reasonable expectation of success. Nothing inventive could be seen in the use of these probes for which sequences and function were already described in the prior art.

Document (17) showed that a cDNA encoding the EGV from Humicola did not hybridize (using a very particular set of stringency conditions, which were not representative of the ones used in the prior art) to cDNAs encoding the CBHI from different organisms. This evidence was irrelevant since it was obvious to the skilled person that probes based on the known homologous terminal CBDs were a better choice than a sequence encoding the full-length CBHI enzyme or the catalytic domain thereof.

Document (16) was also not suited to prove that the claimed enzymes had unexpectedly advantageous properties since deficiencies could be identified in the protocol used to obtain the results described. In particular, only a single, very specific type of substrate was used, there was no indication of the kinetics of the reaction and the enzymes compared therein were of different types, namely the endoglucanase I (EGI) from T. reesei and the EGV from H. insolens. These enzymes had different specificities, optimal conditions of activity, storage requirements, etc. Moreover, the alleged results obtained with the EGV from H. insolens could not be extrapolated to the other enzymes of claim 1, which comprised hemicellulose-degrading enzymes of a very different nature from that of the EGV (specificity, activity, etc.). In any case, the cloning strategy disclosed in the patent was obvious to the skilled person and thus, the claimed sequences were also obvious to achieve. If a surprising effect was associated to those sequences, then it was only a bonus that could not make them less obvious.

In line with the established case law (T 1329/04, supra), post-published evidence could not serve as the sole basis to establish that a patent solved the problem that it purported to solve. In the present case, the patent did not disclose any activity for the enzymes of claim 1, their structure was incompletely characterised (partial amino acid sequence of only one of their domains, namely the CBD), and no further structural or functional features were described. Since the technical problem addressed by the patent was not credibly solved in the patent per se, document (16) which aimed at demonstrating the advantages of the alleged solution should not be taken into account.

Auxiliary requests 1 and 2

Article 56 EPC; Claim 1

Claim 1 was not restricted to the specific EGV from H. insolens - i.e. comprising a defined catalytic domain, linker B region and CBD sequence - used in document (16). There was no reason nor any evidence on file that allowed an extrapolation of the results obtained with EGV (alleged high activity) to other enzymes that shared only a common CBD sequence. Thus, even if the results obtained in document (16) in relation to EGV were considered surprising - which they should not be - they could not serve as a basis to acknowledge inventive step to the other enzymes.

XV. The appellant (patentee) requested that the decision under appeal be set aside and that the patent be maintained in amended form on the basis of the main request as filed with the grounds of appeal, alternatively on the basis of the first or second auxiliary request as filed during oral proceedings.

XVI. The respondent (opponent) requested that the appeal be dismissed.