`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`______________________
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`______________________
`
`DANISCO US INC. and DUPONT NUTRITION BIOSCIENCES ApS,
`Petitioners
`v.
`NOVOZYMES A/S
`Patent Owner
`______________________
`
`Case No. IPR2021-00189
`Patent No. 10,555,541
`______________________
`
`DECLARATION OF DOUGLAS S. CLARK, Ph.D.
`
`
`
`
`

`

`
`
`Contents
`I.
`INTRODUCTION ........................................................................................... 3
`QUALIFICATIONS ........................................................................................ 3
`II.
`III. BACKGROUND AND STATE OF THE ART .............................................. 7
`A.
`Lactose Hydrolysis and Transgalactosylation ....................................... 7
`B.
`Prior Art β-Galactosidase Enzymes from Bifidobacterium
`bifidum ................................................................................................. 10
`1. Moller and Jorgensen ................................................................ 10
`2.
`Larsen and BIF917 .................................................................... 11
`3.
`BIF917 Variants ........................................................................ 14
`IV. The ’541 Patent .............................................................................................. 17
`V.
`Level of Ordinary Skill in the Art ................................................................. 19
`VI. Claim Construction ........................................................................................ 19
`VII. Legal Principles Applied ............................................................................... 20
`VIII. Analysis of the ’508 Application’s Specification .......................................... 21
`A.
`The ’508 Application Describes Only Lactase Enzymes, Not
`Transgalactosylating Enzymes ............................................................ 21
`The ’508 Application Does Not Describe any Fragments of a
`28-979 Amino Acid Fragment of SEQ ID NO: 1 ............................... 28
`The ’508 Application Does Not Describe any Enzymes Having
`Primarily Transgalactosylating Activity ............................................. 29
`Transgalactosylating Activity Is Not Inherent in the 28-979
`Amino Acid Fragment of SEQ ID NO: 1 or Smaller Fragments
`Thereof ................................................................................................ 31
`Conclusion Regarding the Description of the ’508 Application ......... 32
`E.
`IX. Obviousness Analysis of Claims 1, 3-9, and 11-17 of the ’541 Patent ......... 33
`
`D.
`
`B.
`
`C.
`
`1
`
`

`

`
`
`
`
`
`A.
`
`B.
`
`C.
`
`2.
`
`It Would Have Been Obvious to Make a Fragment of Claim 1
`Based on Larsen .................................................................................. 33
`1.
`Larsen Discloses a BIF917 Enzyme Having
`Transgalactosylating Activity that Is Closely
`Homologous to the Polypeptide Fragments of Claim 1............ 33
`Larsen Discloses Making BIF917 Variants, Including
`Variants with Single Conservative Mutations .......................... 35
`Claims 3-9 and 11-17 Are Also Obvious Based on Larsen ................ 39
`1.
`Claim 3 ...................................................................................... 40
`2.
`Claim 4 ...................................................................................... 40
`3.
`Claim 5 ...................................................................................... 41
`4.
`Claim 6 ...................................................................................... 41
`5.
`Claim 7 ...................................................................................... 42
`6.
`Claim 8 ...................................................................................... 43
`7.
`Claim 9 ...................................................................................... 43
`8.
`Claim 11 .................................................................................... 44
`9.
`Claim 12 .................................................................................... 45
`10. Claim 13 .................................................................................... 46
`11. Claim 14 .................................................................................... 46
`12. Claim 15 .................................................................................... 47
`13. Claim 16 .................................................................................... 47
`14. Claim 17 .................................................................................... 48
`Conclusion Regarding Obviousness ................................................... 49
`
`2
`
`

`

`
`
`I.
`
`INTRODUCTION
`I, Douglas S. Clark, Ph.D., have been retained by Finnegan,
`1.
`
`Henderson, Farabow, Garrett & Dunner, LLP, on behalf of Danisco US Inc. and
`
`DuPont Nutrition Biosciences ApS as an independent expert in the fields of
`
`chemical and biochemical engineering, including the use of enzymes in industrial
`
`processes.
`
`2.
`
`I am being compensated for the time I spend on this matter, but my
`
`compensation is not contingent upon my opinions or the outcome of this or any
`
`other proceeding.
`
`II. QUALIFICATIONS
`I am currently Dean of the College of Chemistry and Professor in the
`3.
`
`Department of Chemical and Biomolecular Engineering at the University of
`
`California, Berkeley. I am also a faculty scientist at Lawrence Berkeley Laboratory,
`
`and I hold the endowed G.N. Lewis Chair.
`
`4.
`
`Prior to my appointment as Dean, I served as Department Chair of
`
`Chemical and Biomolecular Engineering and Executive Associate Dean in the
`
`College of Chemistry at Berkeley. I have been a faculty member at Berkeley since
`
`1986.
`
`3
`
`

`

`
`
`5.
`
`I received a Ph.D. in Chemical Engineering from the California
`
`Institute of Technology in 1983 and a B.S. in Chemistry, summa cum laude, from
`
`the University of Vermont in 1979.
`
`6. My research is in the field of biochemical engineering, with particular
`
`emphasis on enzyme technology, biomaterials, and bioenergy, including the
`
`enzymatic hydrolysis of biomass. My laboratory research involving carbohydrate
`
`hydrolysis began when I was a graduate student in 1980. Since then, of my 275
`
`published papers, many deal with hydrolysis-related enzyme applications and
`
`enzyme engineering, including optimization of enzyme stability, production, and
`
`properties. I have additional experience investigating extremophilic enzymes at
`
`elevated temperatures, engineering improved enzyme thermostability, and the
`
`hydrolysis and conversion of lignocellulose to fermentation products.
`
`7.
`
`Throughout my 36-year career, I have taught undergraduate and
`
`graduate courses in biochemical engineering and laboratory techniques, including
`
`enzyme assays, enzyme immobilization, fermentation, and protein purification. My
`
`teaching experience broadly includes glycoside hydrolases, as well as β-
`
`galactosidases specifically, which belong to glycoside hydrolase families GHF-1,
`
`GHF-2, GHF-35, and GHF-42.
`
`8.
`
`For example, I helped devise and teach a biochemical engineering
`
`laboratory class using β-galactosidase from Escherichia coli as a model enzyme. In
`
`4
`
`

`

`
`
`this laboratory class, I instructed students on determining and comparing the
`
`kinetics of β-galactosidase in soluble and immobilized form. This laboratory class
`
`is typically taken by juniors and seniors in chemical engineering and related fields,
`
`who after the exercises are quite proficient in measuring and analyzing the kinetics
`
`of β-galactosidase.
`
`9. My research experience also includes work with cellulases, another
`
`type of glycoside hydrolase. In a series of research papers, we published several
`
`studies of cellulase function and multiple examples of improving cellulase
`
`properties through protein engineering (e.g., site-directed mutagenesis and directed
`
`evolution via DNA shuffling). These publications describe mechanistic modeling
`
`of cellulase mixtures for degrading cellulose; molecular-level studies of enzyme
`
`binding and binding-site heterogeneity in cellulosic substrates; and bioprospecting
`
`and protein engineering of improved cellulases for process conditions (e.g., greater
`
`thermostability, reduced lignin inhibition, and lower product inhibition). Much of
`
`this work utilized standard mutagenesis procedures and methodologies that were
`
`well known to practitioners in the field.
`
`10.
`
`I have served as the Editor-in-Chief of the journal Biotechnology &
`
`Bioengineering since 1996, and I have served on the editorial board of the journal
`
`Extremophiles since 1996 and on the journal Enzyme and Microbial Technology
`
`from 1994 to 2019. All of these journals publish extensively on industrial
`
`5
`
`

`

`
`
`biocatalysis, enzyme technologies, bioprocessing of starch and other biomass, and
`
`fermentation of the resulting sugars to produce alcohols and other products.
`
`11. My textbook, Biochemical Engineering, contains material on enzyme
`
`kinetics, proteases, enzyme denaturation, fermentation, and bioproduct recovery,
`
`among other topics.
`
`12.
`
`I have received a number of honors and awards for my work. I am a
`
`member of the National Academy of Engineering. I am a Fellow of the American
`
`Association for the Advancement of Science and the American Institute of Medical
`
`and Biomedical Engineers. I received both the Daniel I.C. Wang Award and the
`
`James E. Bailey Award from the Society of Biological Engineering; the Marvin J.
`
`Johnson Award in Microbial and Biochemical Technology from the American
`
`Chemical Society; the Food, Pharmaceutical, and Bioengineering Award of the
`
`American Institute of Chemical Engineers; the Amgen Award in Biochemical
`
`Engineering; the International Enzyme Engineering Award; and the NorCal
`
`Chemical Engineering Award—Industrial Research. I also received the
`
`Departmental Chemical Engineering Teaching Award and the Presidential Young
`
`Investigator Award (National Science Foundation).
`
`13. My professional qualifications are described in further detail in my
`
`curriculum vitae, which is attached as Appendix A.
`
`6
`
`

`

`
`
`III. BACKGROUND AND STATE OF THE ART
`A. Lactose Hydrolysis and Transgalactosylation
`β-Galactosidase enzymes, which are sometimes also referred to as
`14.
`
`lactase enzymes, catalyze the hydrolysis of β-galactoside carbohydrates into
`
`monosaccharides through the breaking of a glycosidic bond. Ex. 1004, 647. These
`
`enzymes are often used to hydrolyze the sugar lactose naturally present in milk,
`
`making low-lactose or lactose-free dairy products suitable for consumption by
`
`individuals unable to properly digest dairy products. During lactose hydrolysis, β-
`
`galactosidase cleaves lactose into equal amounts of two products, glucose and
`
`galactose.
`
`15. Some β-galactosidase enzymes can also convert lactose into
`
`galactooligosaccharides through a different reaction known as transgalactosylation.
`
`Ex. 1005, 2276; Ex. 1006, 348-49, 352; Ex. 1003, ¶¶[0003], [0071]. During
`
`transgalactosylation, the enzyme breaks lactose into glucose and galactose and
`
`transfers galactose to an accepting alcohol group of another carbohydrate (e.g.,
`
`glucose, galactose, lactose, or galactose-containing oligosaccharides), building
`
`carbohydrate chains known as galactooligosaccharides (“GOS”). Ex. 1004, 651.
`
`The different reaction pathways for lactose hydrolysis and transgalactosylation are
`
`illustrated below. The figure includes a representative GOS structure (having β-
`
`(1,6) linkages within the brackets) in which p represents the degree of
`
`7
`
`

`

`
`
`polymerization (typically ranging from two to eight monomeric units). Ex. 1009,
`
`5811.
`
`Figure 1
`
`
`
`16. The reaction specificity of a particular β-galactosidase enzyme
`
`depends on a number of factors, including enzyme activity, substrate concentration
`
`(lactose), reaction temperature, enzyme source, and the length of reaction. Ex.
`
`1007, 472, 474-75; Ex. 1006, 348-49. High lactose concentrations, for example,
`
`may enhance transgalactosylation due to the increased availability of lactose for
`
`enzymatic transfer of galactose to build galactooligosaccharides. Ex. 1007, 472,
`
`474-75.
`
`17. Various analytical methods were available to measure enzymatic
`
`activity of β-galactosidases. High-performance liquid chromatography (HPLC), for
`
`example, was a well-known analytical technique for measuring β-galactosidase
`
`8
`
`

`

`
`
`activity by using lactose as a substrate and measuring the resulting concentrations
`
`of glucose, galactose, and/or GOS. Ex. 1007, 478.
`
`18.
`
`If equal levels of galactose and glucose are produced, this indicates
`
`that no transferase (i.e., transgalactosylating) activity occurred. Ex. 1007, 473. If
`
`less galactose than glucose is produced, this indicates that GOS have been
`
`produced and transgalactosylation occurred. See, e.g., Ex. 1001, 18:17-23.
`
`19.
`
`β-Galactosidases having transgalactosylating activity and those having
`
`primarily lactase activity were both recognized as useful, but for different types of
`
`applications. Transgalactosylases, for example, were recognized as useful for
`
`producing GOS in dairy products. Ex. 1007, 472. GOS are non-digestible
`
`prebiotics that promote proliferation of microorganisms, such as healthy bacteria in
`
`yogurt, that can improve digestion and promote growth of intestinal microflora.
`
`Ex. 1005, 2276; Ex. 1007, 471-72.
`
`20. Hydrolyzing β-galactosidases, by contrast, were used in industrial
`
`processes for making low-lactose or lactose-free products. In these processes, GOS
`
`potentially produced through a transgalactosylation side-reaction were recognized
`
`as undesirable byproducts. As one article summarized, “[i]ndustrial processes
`
`aimed at producing low-lactose or lactose-free items are concerned with
`
`undesirable GOS byproducts, for fear of unknown side effects that may stimulate
`
`9
`
`

`

`
`
`symptoms of lactose intolerance.” Ex. 1009, 5811. Thus, highly efficient
`
`hydrolyzing enzymes were used for producing low-lactose or lactose-free products.
`
`21. Producing lactose-free products through hydrolysis or producing GOS
`
`through transgalactosylation were also recognized as divergent processes with
`
`different aims. Conducting lactose hydrolysis with efficient lactase enzymes has
`
`the intended goal of complete lactose conversion to make lactose-free products.
`
`Producing GOS, in contrast, typically requires proceeding to only partial
`
`conversion, leaving significant quantities of residual lactose. Ex. 1006, 347
`
`(Figure 2). As researchers recognized, transgalactosylating enzymes can break
`
`down both lactose and GOS. Ex. 1006, 347; Ex. 1009, 5817. Thus, running a
`
`transgalactosylation reaction too far, resulting in low levels of lactose, makes GOS
`
`comparatively more available as a substrate and results in counterproductive
`
`breakdown of GOS. Ex. 1006, 347; Ex. 1009, 5817.
`
`Prior Art β-Galactosidase Enzymes from Bifidobacterium bifidum
`B.
`1. Moller and Jorgensen
`22. Moller, published in 2001, states that some β-galactosidases may
`
`catalyze the formation of GOS through transgalactosylation in addition to their
`
`normal hydrolysis activity. Ex. 1005, 2276. Moller reports the screening and
`
`identification of one such β-galactosidase from Bifidobacterium bifidum
`
`DSM20215, identified as BIF3. Ex. 1005, 2276.
`
`10
`
`

`

`
`
`23. The same research group conducted additional studies on this BIF3
`
`enzyme, published by Jorgensen et al. in 2001. Ex. 1004. To investigate the
`
`functional importance of the C-terminal portion of BIF3, Jorgensen constructed
`
`four deletion mutants using standard restriction enzymes to truncate the β-
`
`galactosidase gene from the C-terminal end of the coding region. Ex. 1004, 648.
`
`24.
`
`Jorgensen showed that one of its truncated β-galactosidases, BIF3-d3,
`
`produced significantly more GOS compared to the full-length BIF3 enzyme. Ex.
`
`1004, 651, Figs. 3 and 4. The shortest, BIF3-d4, did not exhibit any appreciable
`
`enzymatic activity (0.1%). Ex. 1004, 648-49.
`
`25. Much remained unknown about BIF3 and its activity. For example,
`
`Jorgensen emphasized that the three-dimensional structure of BIF3 was unknown,
`
`and therefore it would be difficult to predict advantageous, site-specific changes in
`
`the enzyme. Ex. 1004, 651. Jorgensen also stated that BIF3’s mechanism of action
`
`was unknown and remained to be elucidated. Ex. 1004, 652. The authors
`
`speculated that binding between enzyme and acceptor may be a requirement for the
`
`BIF3-d3-catalysed reaction. Ex. 1004, 652.
`
`2.
`26.
`
`Larsen and BIF917
`In 2015, fourteen years after Jorgensen, Larsen published further
`
`studies on β-galactosidase truncation mutants. Ex. 1003.
`
`11
`
`

`

`
`
`27. Larsen investigated β-galactosidase fragments across a range of eight
`
`fragment lengths. Starting with the full-length β-galactosidase from
`
`Bifidobacterium bifidum DSM20215 (1752 amino acid residues), Larsen generated
`
`eight fragments between 887 amino acids and 1448 amino acids in length.
`
`• BIF917 (887 amino acids)
`
`• BIF995 (965 amino acids)
`
`• BIF1068 (1038 amino acids)
`
`• BIF1172 (1142 amino acids)
`
`• BIF1241 (1211 amino acids)
`
`• BIF1326 (1296 amino acids)
`
`• BIF1400 (1370 amino acids)
`
`• BIF1478 (1448 amino acids)
`
`Ex. 1003, ¶¶[0469], [0042]-[0047], [0063], [0205], [0440].
`
`28. Larsen published the amino acid sequence of BIF917, which is
`
`disclosed as SEQ ID NO: 1. Ex. 1003, ¶[0042]. Larsen also determined the active
`
`site (also referred to as a “catalytic core”) of BIF917 and the other truncation
`
`mutants, which is disclosed as SEQ ID NO: 7. Ex. 1003, ¶¶[ [0048], [0176]. Within
`
`the sequence of BIF917, shown below, the amino acid residues comprising the
`
`catalytic core are shown in purple, spanning residues 29-481. The N-terminal
`
`region outside of the catalytic core is shown in blue.
`
`12
`
`

`

`
`
`Figure 2
`
`
`
`29. Larsen evaluated its enzyme fragments in milk-based applications. In
`
`Example 2, Larsen demonstrated that BIF917 and BIF995 produced significantly
`
`more GOS (~1.2% w/v) than a longer fragment, BIF1326 (below 0.1% w/v), when
`
`tested at a lactose concentration of 5.5% (w/v). Ex. 1003, ¶¶[0472]-[0474]; Fig. 4.
`
`30.
`
`In Example 4, Larsen evaluated BIF917 and BIF995 in yogurt
`
`mixtures. Ex. 1003, ¶¶[0479]-[0482]. Larsen reported that increasing doses of
`
`BIF917 or BIF995 from 10 ppm to 40 ppm led to increased GOS content and
`
`decreased levels of lactose (measured as DP2 saccharides) in the final yogurt. Ex.
`
`1003, ¶[0482], Table 3. When the initial lactose concentration was increased from
`
`5.5% to 7.5% w/v, BIF917 produced a final GOS concentration of 2.954% and
`
`2.662% at enzyme doses of 50 ppm and 25 ppm, respectively, whereas the control
`
`β-galactosidase produced only 0.355% GOS. Ex. 1003, ¶¶[0483]-[0490], Table 5.
`
`Larsen also demonstrated that different yogurt cultures, temperatures, and
`
`acidification profiles had no significant effect on the final GOS yield. Ex. 1003,
`
`¶[0504], Table 7. Larsen concluded that BIF917 exhibited “robust”
`
`13
`
`

`

`
`
`transgalactosylating activity in yogurt under a range of culture and temperature
`
`conditions. Ex. 1003, ¶[0504].
`
`31. Larsen also described its enzymes’ activity in terms of a ratio of
`
`transgalactosylation activity. Ex. 1003, ¶[0471]. Larsen states that ratios above
`
`100% indicate that an enzyme is predominantly transgalactosylating, and ratios
`
`below 100% indicate that an enzyme is predominantly hydrolytic. Ex. 1003,
`
`¶[0471]. Larsen reports that, based on its experiments, BIF917 exhibited a ratio of
`
`transgalactosylation activity of around 250%. Ex. 1003, ¶[0471]. The next-largest
`
`enzyme, BIF995 (965 amino acids long), exhibited a similarly high ratio of
`
`transgalactosylation activity. Ex. 1003, ¶[0471].
`
`32. Larsen reported that the enzymes it made having a length of 1241
`
`residues or less showed a ratio of transgalactosylation activity above 100%,
`
`indicating that they were predominantly transgalactosylating. Ex. 1003, ¶¶[0469]-
`
`[0471] (Example 1). For the enzymes Larsen made that were longer than 1241
`
`residues, Larsen reported that the enzymes were predominantly hydrolytic (i.e.,
`
`had lactase activity). Ex. 1003, ¶[0471], Fig. 3.
`
`BIF917 Variants
`3.
`33. Larsen also discloses variants of BIF917 having as little as one amino
`
`acid substitution. Ex. 1003, ¶[0210]. Larsen states that such a variant may have
`
`between one and ten substitutions within the amino acid sequence of SEQ ID NO:
`
`14
`
`

`

`
`
`1 (i.e., BIF917). Ex. 1003, ¶[0219]. Larsen further states that a single amino acid
`
`substitution should be “in the N-terminal end” of SEQ ID NO: 1. Ex. 1003,
`
`¶[0219].
`
`34. Larsen encourages making BIF917 variants with a single,
`
`conservative mutation. Ex. 1003, ¶¶[0198]-[0200]. A conservative mutation is the
`
`replacement of one amino acid with another amino acid with similar properties
`
`(e.g., charge, hydrophobicity, and size). Such a mutation is “conservative” because
`
`it is expected to maintain the specific activity and structure of the enzyme. Ex.
`
`1010, 9:36-50.
`
`35. Larsen states that BIF917 variants with as little as one conservative
`
`mutation would be expected to maintain transgalactosylating activity. Larsen, for
`
`example, discloses that the single mutation to SEQ ID NO: 1 may be a
`
`conservative substitution that produces a “silent change” and results in a
`
`functionally equivalent enzyme. Ex. 1003, ¶¶[0198], [0200]. Larsen also states that
`
`deliberate amino acid substitutions may be made based on similarity of one amino
`
`acid to another. Ex. 1003, ¶¶[0198]. Larsen proposes making like-for-like
`
`substitutions of amino acid residues within the sequence of BIF917. Ex. 1003,
`
`[0199].
`
`15
`
`

`

`
`
`36. Small groups of amino acids that have similar properties, and
`
`therefore may be substituted to conserve functional properties, have long been
`
`known, including, for example:
`
`• Basic amino acids: arginine, lysine, and histidine
`
`• Acidic amino acids: glutamic acid and aspartic acid
`
`• Polar amino acids: glutamine, asparagine, serine, and threonine
`
`• Hydrophobic amino acids: leucine, isoleucine, and valine
`
`• Aliphatic amino acids: alanine, valine, leucine, and isoleucine
`
`• Aromatic amino acids: phenylalanine, tryptophan, and tyrosine
`
`Ex. 1010, 9:36-50; Ex. 1003, ¶[0200].
`
`37. As Larsen discloses, for example, mutating valine to alanine, leucine,
`
`or isoleucine would be a conservative substitution based on the similar properties
`
`of the small group of aliphatic amino acids. Ex. 1003, ¶¶[0200], [0321]. Larsen
`
`discloses that such closely homologous variants will retain the same active sites
`
`and should retain or enhance the functional transgalactosylating activity compared
`
`to a polypeptide of SEQ ID NO: 1 (i.e., BIF917). Ex. 1003, ¶¶[0195]-[0196].
`
`38.
`
`It was also routine in the art to identify and alter enzymes at one or
`
`more amino acid residues to produce homologous variants with similar or
`
`enhanced functional properties. Strategies such as site-directed mutagenesis,
`
`alanine-scanning mutagenesis, recombination, and shuffling were well-known
`
`16
`
`

`

`
`
`techniques for identifying and creating single or multiple amino acid substitutions.
`
`Ex. 1010, 9:22-10:44. Those techniques were often performed with high-
`
`throughput, automated screening methods, which allowed rapid evaluation of
`
`individual mutations across a wide span of polypeptide variants. Ex. 1010, 9:22-
`
`10:44. Accordingly, one of ordinary skill in the art would have been ready capable
`
`of following Larsen’s guidance to produce polypeptide variants of BIF917 while
`
`maintaining transgalactosylating activity.
`
`IV. The ’541 Patent
`39. The ’541 patent is titled “Method for Producing a Dairy Product.”
`
`Ex. 1001, title. It states that the application for the ’541 patent was filed April 10,
`
`2019, and is related to the ’107 patent filed February 15, 2017. The abstract states
`
`that the invention relates to a method for producing a dairy product using an
`
`enzyme having lactase activity. Ex. 1001, abstract. The ’541 patent defines lactase
`
`activity as lactose hydrolyzing activity. Ex. 1001, 11:38-41.
`
`40. The specification of the ’541 patent relates to enzymes having lactase
`
`activity, indicating that transgalatosylating activity is undesired. The ’541 patent
`
`states in the background section, for example, that a lactase should hydrolyze
`
`lactose efficiently and optimally allow for almost complete lactose hydrolysis. Ex.
`
`1001, 2:25-29. The ’541 patent also states that its lactases should be used to make
`
`dairy products that are not enriched by galactooligosaccharides, which are the
`
`17
`
`

`

`
`
`product of transgalactosylation. Ex. 1001, 5:49-51. The ’541 patent also states that
`
`it is preferred for a lactase to produce equal amounts of glucose and galactose—
`
`i.e., complete hydrolysis with no transgalactosylation. Ex. 1001, 15:48-51. The
`
`examples of the ’541 patent consistently report the absence of any
`
`transgalactosylating activity. Ex. 1001, 16:37-18:23, 24:1-57. I analyze the
`
`specification of the ’508 application, which is substantively identical to the
`
`specification of the ’541 patent, in detail below in Section VIII.
`
`41. The claims of the ’541 patent describe enzymes having
`
`transgalactosylating activity. Claim 1 describes a polypeptide having
`
`transgalactosylating activity, which is a truncated polypeptide consisting of the
`
`amino acid sequence of amino acids 28-979 of SEQ ID NO: 1 or is a fragment
`
`thereof having transgalactosylating activity, wherein the polypeptide is isolated.
`
`Claim 11 describes a fragment having an amino acid sequence which is at least
`
`98% identical to amino acids 28-979 of SEQ ID NO: 1, wherein the fragment
`
`consists of at most 952 amino acid residues and has transgalactosylating activity,
`
`wherein the polypeptide is isolated.
`
`42. Claims 3-9 and 12-17 describe additional features of the enzymes
`
`having transgalactosylating activity or conventional applications for such enzymes:
`
`• a fragment of amino acids 28-979 of SEQ ID NO: 1 (claim 3);
`
`• a composition comprising the polypeptide (claims 4 and 12);
`
`18
`
`

`

`
`
`• a food composition (claims 5 and 13);
`
`• a dairy product composition (claims 6 and 14);
`
`• a method for producing a food product, comprising treating a
`
`substrate comprising lactose with the polypeptide (claims 7 and 15);
`
`• a method for producing a dairy product, comprising treating a milk-
`
`based substrate comprising lactose with the polypeptide (claims 8 and
`
`16); and
`
`• a method for producing galactooligosaccharides comprising
`
`contacting the polypeptide with a milk-based substrate comprising
`
`lactose (claims 9 and 17).
`
`V. Level of Ordinary Skill in the Art
`I believe that a person of ordinary skill in the art as of the May 25,
`43.
`
`2010 filing date of the ’508 application and the April 10, 2019 filing date of the
`
`’541 patent would have had a Ph.D. or master’s degree in chemistry, chemical
`
`engineering, biochemical engineering, molecular biology, or a related technical
`
`field, with several years of experience studying, developing, or using enzymes
`
`such as β-galactosidases in industrial processes.
`
`VI. Claim Construction
`44. The term “transgalactosylating activity” is used in claims 1 and 11 of
`
`the ’541 patent. A person of ordinary skill in the art would have understood that
`
`19
`
`

`

`
`
`transgalactosylating activity refers to the transfer of a galactose moiety to a
`
`molecule other than water to produce galactooligosaccharides. Ex. 1004, 651; Ex.
`
`1003, ¶[0068]. This common understanding of transgalactosylating activity is
`
`consistent with the ’541 patent specification, which states that transgalactosylation
`
`produces galactooligosaccharides, resulting in lower galactose concentrations
`
`compared to glucose. Ex. 1001, 16:37-18:23, 24:1-57.
`
`45. The claims do not describe any particular level or ratio of
`
`transgalactosylating activity. To a person of ordinary skill, when a single activity is
`
`used to characterize an enzyme, such as transgalactosylating activity in claims 1
`
`and 11, it typically refers to the enzyme’s primary activity. Consistent with that
`
`understanding, Novozymes stated during prosecution that its claims covered
`
`enzymes having a ratio of transgalactosylation activity above 150%. Ex. 1012, 5.
`
`Claims 1 and 11 therefore encompass enzymes having transgalactosylating activity
`
`as their primary activity, such as those having ratios of transgalactosylating activity
`
`to lactase activity above 1:1.
`
`VII. Legal Principles Applied
`I have been asked to analyze whether Novozymes’ ’508 application,
`46.
`
`filed May 25, 2010, describes the enzymes having transgalactosylating activity of
`
`claims 1, 3-9, and 11-17 of the ’541 patent. In conducting this analysis, I have
`
`analyzed the specification of the ’508 application from the perspective of a person
`
`20
`
`

`

`
`
`of ordinary skill in the art as of May 25, 2010 to evaluate whether the inventors
`
`actually invented the subject matter of the ’541 patent claims.
`
`47.
`
`I have also been asked to analyze whether claims 1, 3-9, and 11-17 of
`
`the ’541 patent would have been obvious to a person of ordinary skill in the art as
`
`of the April 10, 2019 filing date. In conducting this analysis, I have considered the
`
`scope and content of the prior art, any differences between the claims and the prior
`
`art, and the level of ordinary skill in the art. I have also considered whether a
`
`person of ordinary skill in the art would have had reasons or motivations to modify
`
`the prior art with a reasonable expectation of successfully achieving the claimed
`
`invention.
`
`VIII. Analysis of the ’508 Application’s Specification
`A. The ’508 Application Describes Only Lactase Enzymes, Not
`Transgalactosylating Enzymes
`48. The ’508 application does not describe any transgalactosylating
`
`enzymes as its invention. Rather, the entire application is directed solely to lactase
`
`enzymes. The examples of the ’508 application, for instance, evaluated only a
`
`single lactase of amino acids 28-1331 of SEQ ID NO: 2 from Bifidobacterium
`
`bifidum. Ex. 1015, 22:28-33. The examples reported that it had no
`
`transgalactosylating activity.
`
`49. Example 2 evaluated the SEQ ID NO: 2 lactase in a milk solution
`
`having 5% lactose at 37ºC, using a previously known commercial lactase
`
`21
`
`

`

`
`
`(Lactozym) as a positive control. Ex. 1015, 22:25-31. Example 2 reports lactose,
`
`glucose, and galactose concentrations measured by HPLC over four hours of
`
`sampling. It states that no transferase activity was observed when using the
`
`Bifidobacterium lactase. Ex. 1015, 23:10-12. Transferase activity is another term
`
`for transgalactosylating activity. Example 2 further reports that equal amounts of
`
`glucose and galactose were produced, which is the result of lactase (hydrolysis)
`
`activity with no transgalactosylation. Ex. 1015, 23:11-12. Example 2 confirmed its
`
`result by reference to the positive control, reporting that Lactozym produced less
`
`galactose than glucose, showing that galactooligosaccharides were produced.
`
`Ex. 1015, 23:10-15.
`
`50. Example 3 evaluated the SEQ ID NO: 2 lactase with a different
`
`substrate: a whey permeate solution at 37ºC. It states that the whey permeate
`
`solution used was primarily lactose and ions dissolved in water. Ex. 1015,
`
`23:17-29. Example 3 reports lactose, glucose, and galactose concentrations
`
`measured by HPLC over five and a half hours of sampling. It again reports that the
`
`galactose and glucose levels were equal, which is the result of lactase (hydrolysis)
`
`activity with no transgalactosylation. Ex. 1015, 24:14-15. Example 3 concludes
`
`that no or very little galactooligosaccharides were produced. Ex. 1015, 24:13-17.
`
`Example 3 confirmed its result by reference to the positive control, reporting that
`
`22
`
`

`

`
`
`Lactozym produced less galactose than glucose, showing production of
`
`galactooligosaccharides. Ex. 1015, 24:13-17
`
`51. Example 10 evaluated the SEQ ID NO: 2 lactase in a milk solution
`
`having 5% lactose at 5ºC. Ex. 1015, 32:6-21. Example 10 reports lactose, glucose,
`
`and galactose concentrations measured by HPLC over forty-eight hours of
`
`sampling. It again reports equal production of glucose and galactose, i.e., the result
`
`of lactose hydrolysis with no transgalactosylation. Ex. 1015, 33:3-8. Example 10
`
`similarly concludes that no transferase activity was observed when using the
`
`Bifidobacterium lactase. Ex. 1015, 33:3-8. By compa