Trials@uspto.gov
`Tel: 571-272-7822
`
`Paper 13
`Entered: March 9, 2018
`
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
`
`COHERUS BIOSCIENCES, INC.,
`Petitioner,
`v.
`HOFFMAN-LaROCHE INC.,
`Patent Owner.
`
`Case IPR2017-01916
`Patent 8,163,522 B1
`
`
`Before SUSAN L. C. MITCHELL, TINA E. HULSE, and
`WESLEY B. DERRICK, Administrative Patent Judges.
`
`MITCHELL, Administrative Patent Judge.
`
`
`
`
`DECISION
`Denying Institution of Inter Partes Review
`37 C.F.R. § 42.108
`
`
`
`
`
`
`
`
`
`
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`
`I. INTRODUCTION
`
`A. Background
`Petitioner Coherus Biosciences, Inc. (“Petitioner”) filed a petition
`(Paper 1, “Pet.”) to institute an inter partes review of claims 1–10 (the
`“challenged claims”) of U.S. Patent No. 8,163,522 B1 (Exhibit 1001, “the
`’522 patent”). See 35 U.S.C. §§ 311–319. Patent Owner Hoffman-LaRoche
`Inc. (“Patent Owner”), filed a Preliminary Response. Paper 9 (“Prelim.
`Resp.”).
`
`We have authority to determine whether to institute an inter partes
`review. See 35 U.S.C. § 314(b); 37 C.F.R. 42.4(a). To institute an inter
`partes review, we must determine that the information presented in the
`Petition shows “a reasonable likelihood that the petitioner would prevail
`with respect to at least 1 of the claims challenged in the petition.” 35 U.S.C.
`§ 314(a). For the reasons set forth below, we conclude that Petitioner has
`not established a reasonable likelihood that it would prevail in showing the
`unpatentability of any challenged claim of the ’522 patent. Therefore, we do
`not institute an inter partes review for any challenged claim of the ’522
`patent.
`
`B. Related Proceedings
`The parties identify two court proceedings involving the ’522 patent,
`one of which has been terminated and one that is ongoing: Sandoz Inc. v.
`Amgen Inc., 773 F.3d 1274 (Fed. Cir. 2014) (terminated) and Immunex
`Corp. v. Sandoz Inc., Case No. 2:16-cv-01118-CCC-JBC (D.N.J.) (pending).
`Pet. 7; Paper 8, 2.
`The parties also identify a previously filed request for inter partes
`review of the ’522 patent that was not instituted: Coalition for Affordable
`
`2
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`Drugs V LLC v. Hoffman-LaRoche Inc., Case IPR2015-01792 (PTAB) (“the
`1792 IPR”). Pet. 7, Paper 8, 2; Ex. 1010. Petitioner has also filed a request
`for inter partes review of related U.S. Patent No. 8,063,182 B1 (“the ’182
`patent”), Case IPR2017-02066. Paper 8, 2.
`
`C. The ’522 Patent (Ex. 1001)
`The ’522 patent is directed, in part, to polynucleotides encoding the
`extracellular region of an insoluble human TNF receptor (also, “TNF-R”)
`described by an apparent molecular weight and as containing particular
`amino acid sequences in addition to all domains of the constant region of a
`human IgG1 immunoglobulin heavy chain except the first domain of the
`heavy chain constant region. Ex. 1001, Abs., 2:26–49. The ’522 patent also
`addresses methods for culturing a host cell comprising the polynucleotide
`and purifying the expression product of the polynucleotide from the cell. Id.
`D. Illustrative Claims
`Claims 1 and 4 are illustrative of the claimed subject matter. Claims 1
`and 4 are reproduced below.
`1. A method comprising the steps of:
`(a) culturing a host cell comprising a polynucleotide, wherein
`the polynucleotide encodes a protein consisting of:
`
`
`(i) the extracellular region of an insoluble human TNF
`receptor, wherein the insoluble human TNF receptor has an
`apparent molecular weight of about 75 kilodaltons as
`determined on a non-reducing SDS-polyacrylamide gel and
`comprises the amino acid sequence
`LPAQVAFXPYAPEPGSTC (SEQ ID NO: 10), and
`
`(ii) all of the domains of the constant region of a human
`IgG immunoglobulin heavy chain other than the first domain of
`said constant region, and
`
`3
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`
`(b) purifying an expression product of the polynucleotide from
`the cell mass or the culture medium.
`
`
`Ex. 1001, 45:45–62.
`
`4. A polynucleotide encoding a protein consisting of:
`
`(a) the extracellular region of an insoluble human TNF receptor,
`wherein the insoluble human TNF receptor (i) has an
`apparent molecular weight of about 75 kilodaltons as
`determined on a non-reducing SDS-polyacrylamide gel and (ii)
`comprises the amino acid sequence
`LPAQVAFXPYAPEPGSTC (SEQ ID NO: 10), and
`
`
`
`(b) all of the domains of the constant region of a human IgG1
`immunoglobulin heavy chain other than the first domain of said
`constant region.
`
`Id. at 46:44–55.
`
`E. The Asserted Grounds of Unpatentability
`Petitioner contends that the challenged claims 1–10 are unpatentable
`under 35 U.S.C. § 103(a) based on the following grounds. Pet. 9.
`
`References
`Watson1 and Smith2
`Smith, Watson, and
`Zettlmeissl3
`
`Statutory Basis
`§ 103
`§ 103
`
`Claims Challenged
`1–10
`1–10
`
`
`1 Watson et al., A Homing Receptor–IgG Chimera as a Probe for Adhesive
`Ligands of Lymph Node High Endothelial Venules, 110 J. CELL BIOL. 2221–
`29 (June 1990) (Ex. 1003).
`2 Smith et al., U.S. Patent No. 5,395,760, issued March 7, 1995 (Ex. 1004).
`3 Zettlmeissl et al., Expression and Characterization of Human CD4:
`Immunoglobulin Fusion Proteins, 9 DNA & CELL BIOLOGY 347–53 (June
`1990) (Ex. 1005).
`
`4
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`Petitioner supports the Petition with the testimony of Dennis R.
`Burton, Ph.D. (Ex. 1002).
`
`II. ANALYSIS
`A. Application of 35 U.S.C. § 325(d) or 35 U.S.C. § 314(a)
`Patent Owner asks that we use our discretion under 35 U.S.C.
`§§ 325(d) or 314(a) to deny inter partes review in this case. Prelim. Resp.
`21–31. Specifically, Patent Owner asserts that we should exercise such
`discretion because Smith, used in both grounds in this case, was considered
`in the 1792 IPR and during examination of the ’522 patent, Watson
`describes the same fusion protein as described in references used in the 1792
`IPR challenge and by the Examiner during prosecution, and Zettlmeissl
`discloses the same fusion protein constructs described in a reference
`considered in the 1792 IPR. Id. at 27–28.
`Petitioner asserts that its Petition differs from the petition in the 1792
`IPR because “Watson and Zettlmeissl both provide a clear and compelling
`reason why a POSA would have specifically selected a fusion protein
`incorporating the hinge-CH2-CH3 region of an IgG.” Pet. 17–18.
`Specifically, Petitioner argues that Zettlmeissl reports poor expression for
`fusion proteins with CH1 domains and excellent expression for a receptor
`protein that is joined to the hinge-CH2-CH3 region of human IgG1. Id. at
`18. Likewise, Petitioner asserts that “Watson identifies only one location as
`optimal for fusion of a receptor protein to the immunoglobulin.” Id.
`Instead of analyzing whether there are differences between the art
`asserted in this Petition and that discussed during prosecution of the ’522
`patent or the previous Petition in the 1792 IPR, we find it more efficient to
`resolve our decision on institution on the merits presented in the Petition.
`
`5
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`Because we find that Petitioner has no reasonable likelihood of prevailing on
`either ground presented in the Petition and, therefore, deny institution, we
`decline to exercise our discretion under either 35 U.S.C. § 314(a) or
`35 U.S.C. § 325(d) to deny the Petition.
`
`B. Level of Skill in the Art
`Petitioner states that one of ordinary skill in the art in the relevant
`field of recombinant DNA processes for the production, isolation, and use of
`chimeric proteins “would have held an advanced degree, such as a Ph.D., in
`molecular biology, biochemistry, cell biology, molecular genetics, or a
`related field, and would have experience using recombinant DNA processes
`to construct chimeric proteins, as well as experience using techniques for the
`expression, isolation, and purification of proteins.” Pet. 19–20 (citing
`Ex. 1002 ¶ 30). Patent Owner’s declarant essentially agrees with this
`definition, but states that he “disagree[s] that the Ordinary Artisan would
`have had much experience constructing chimeric proteins, as this was a
`relatively new development in August 1990.” Ex. 2001 ¶ 31.
`We do not discern an appreciable difference in the parties’ respective
`definitions of the level of ordinary skill. Moreover, we note that Patent
`Owner does not contest the level of skill in the art in its Preliminary
`Response. Accordingly, we adopt Petitioner’s uncontested definition of the
`level of ordinary skill mindful that the experience using recombinant DNA
`processes to construct chimeric proteins may have been somewhat limited.
`Ex. 2001 ¶ 31. We further note that the prior art itself demonstrates the level
`of skill in the art at the time of the invention including the level of
`experience using recombinant DNA techniques to construct chimeric
`proteins. See Okajima v. Bourdeau, 261 F.3d 1350, 1355 (Fed. Cir. 2001)
`
`6
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`(explaining that “specific findings on the level of skill in the art . . . [are not
`required] ‘where the prior art itself reflects an appropriate level and a need
`for testimony is not shown’”) (quoting Litton Indus. Prods., Inc. v. Solid
`State Sys. Corp., 755 F.2d 158, 163 (Fed. Cir. 1985)).
`C. Claim Interpretation
`In an inter partes review, claim terms in an unexpired patent are given
`their broadest reasonable construction in light of the specification of the
`patent in which they appear. 37 C.F.R. § 42.100(b); Cuozzo Speed Techs.,
`LLC v. Lee, 136 S. Ct. 2131, 2142 (2016) (affirming applicability of
`broadest reasonable construction standard to inter partes review
`proceedings). Claim terms are given their ordinary and customary meaning
`as would be understood by one of ordinary skill in the art in the context of
`the entire disclosure. In re Translogic Tech., Inc., 504 F.3d 1249, 1257
`(Fed. Cir. 2007). An inventor may rebut that presumption by providing a
`definition of the term in the specification with reasonable clarity,
`deliberateness, and precision. In re Paulsen, 30 F.3d 1475, 1480 (Fed. Cir.
`1994). In the absence of such a definition, limitations are not to be read
`from the specification into the claims. In re Van Geuns, 988 F.2d 1181,
`1184 (Fed. Cir. 1993).
`Each independent claim of the ’522 patent requires that the protein
`encoded includes “all of the domains of the constant region of a human IgG
`[or IgG1] immunoglobulin heavy chain other than the first domain of said
`constant region.” See Ex. 1001, 45:58–60, 46: 53–55, 47:1–3 (claims 1, 4,
`and 7) (emphasis added). Petitioner relies on the Board’s previous
`interpretation in the 1792 IPR of this claim term. Pet. 20–21. The Board
`had previously interpreted this claim phrase to mean “-hinge-CH2-CH3
`
`7
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`region of a human IgG (or IgG1 as appropriate to the requirements of a
`particular claim) immunoglobulin heavy chain.” Id.; Ex. 1010, 7.4 As
`Patent Owner notes, Petitioner relies on these previous constructions without
`further comment. See Prelim. Resp. 13 (citing Pet. 20–21; Ex. 1002 ¶¶ 33–
`34).5
`
`The Petition does not otherwise elaborate on the meaning of the
`phrase, or the import of our earlier determination in the 1792 IPR, as to what
`is, in fact, required by the claims. See Ex. 1010, 7; see generally, Pet. The
`Petition does, however, contend that Watson and Zettlmeissl “both reported
`optimal results by employing the identical portion of the IgG heavy chain as
`claimed in the ’522 patent.” Pet. at 5 (citing Ex. 1002 ¶¶ 76–78, 84–86,
`132). The Petition explains that both references “report[] that receptor:IgG
`hinge fusion proteins are most ‘efficiently synthesized’ when the light chain
`and CH1 domain are deleted, so that the receptor is attached directly to the
`hinge-CH2-CH3 region of an IgG antibody’s heavy chain.” Id. at 4–5
`(citing Ex. 1002 ¶¶ 151–158; Ex. 1003, 2224; Ex. 1005, 347).
`Patent Owner contends that the claims require the proteins to include
`the complete hinge-CH2-CH3 region of the heavy chain, that is, “all of the
`domains of the constant region . . . other than the first domain [CH1] of said
`
`4 In the 1792 IPR we found that two other claim terms, “TNF receptor” and
`“about,” did not require express construction. Ex. 1010, 5, 7. Petitioner
`offered no express construction of either term in this proceeding and asserted
`that each term should be given its ordinary meaning. Pet. 20.
`5 Petitioner requested authorization to file a reply brief to address whether
`the Board’s prior construction of “all of the domains of the constant region .
`. . other than the first domain of said constant region,” to which, Petitioner
`contends, Patent Owner agreed, includes a functional as well as a genetic
`hinge. Paper 12, 10:17–12:3. After extensive discussion by both parties, we
`declined to allow Petitioner an additional brief. Id. at 49.
`
`8
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`constant region.” Prelim. Resp. 7. Patent Owner further contends that the
`Board’s prior construction is wholly consistent with “the claims requir[ing]
`use of the entire hinge-CH2-CH3 region of the IgG/IgG1 heavy chain, not a
`truncated portion of that region.” Id. at 15. Patent Owner notes that the
`claims were distinguished during prosecution “from other fusion proteins
`having ‘only a portion of a hinge domain’” and adds, with emphasis, that as
`“noted during the CFAD-IPR [1792 IPR] . . . the claims ‘were drafted to
`exclude other p75 TNFR/IgG fusion proteins (such as Delta 57 and Protein
`3.5D) that contained only a portion of the hinge domain and did not display
`the unexpected properties.” Id. at 18–19 (citing Ex. 2110, 35; Ex. 1008, 40,
`47).
`
`Patent Owner also provides evidence supporting the contention that
`“the plain and ordinary meaning” of the CH1, hinge, CH2, and CH3
`domains of human IgG heavy chains set forth in our prior decision is the
`respective amino acid sequence “encoded by the CH1, hinge, CH2 and CH3
`exons.” Prelim. Resp. 15–17 (citing Ex. 2012, xix; Ex. 2014, Fig. 4;
`Ex. 1050, 4072; Ex. 2001 ¶¶ 62–64).
`We determine that the broadest reasonable interpretation of the phrase
`“all of the domains of the constant region . . . other than the first domain of
`said constant region” means “all of the hinge, CH2, and CH3 domains.”
`That is, all of the constant region forming domains, i.e., CH1, hinge, CH2,
`and CH3 domains, is included except for the first domain. Thus, any protein
`with less than all of the amino acid sequence of the hinge domain of a
`human IgG (or IgG1) immunoglobulin heavy chain, even if functional, falls
`outside the scope of the claims as properly construed, because it omits a
`portion of the constant region forming domains other than the first domain.
`
`9
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`This construction is consistent with statements the applicant made
`during prosecution that a fusion protein that includes only a portion of a
`hinge domain “are missing the first several amino acids of this domain, and
`thus do not comprise ‘all of the domains of the constant region of a human
`immunoglobulin IgG heavy chain other than the first domain.’” Ex. 2110,
`35. Also, there is evidence in the record that the Specification of the ’522
`patent is consistent with this interpretation on the basis that the described
`fusion proteins include all of the amino acid sequence of the heavy chain
`constant region except the first domain. Ex. 2001 ¶¶ 59, 42–45, 46–47
`(describing Example 11 TNFR-based fusion protein); see also Prelim. Resp.
`18–19, 18 n.35 (discussing vectors used in examples in the ’522 patent
`contained exons encoding the full hinge, CH2, and CH3 domains).
`The question remains, however, where in the constant region the
`divide lies between the first domain of the constant region and the hinge
`domain. Because Petitioner fails to answer this question in a consistent
`manner, we determine on this record that Petitioner has not shown
`sufficiently that the claims are unpatentable as obvious.
`
`D. Asserted References
`1. Watson (Ex. 1003)
`Watson reports the “develop[ment] [of] a chimeric protein containing
`the murine [pln homing receptor] and the hinge and constant regions of the
`human immunoglobulin heavy chains . . . thus, converting the pln HR into a
`monoclonal antibody-like molecule.” Ex. 1003, 2222. Watson describes a
`“truncated [murine homing receptor] protein [] joined to a human heavy
`chain gamma-1 region immediately NH2-terminal to the hinge domain (H)
`such that this chimera contains the two cysteine residues (C) of the hinge
`
`10
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`responsible for immunoglobulin dimerization as well as the CH2 and CH3
`constant regions.” Id. at 2223, Fig. 1. Watson describes data “indicating
`that the hinge region was fully functional in this chimera.” Id. at 2224.
`Watson does not otherwise define the hinge domain, its bounds, or sequence,
`but refers to published work by Capon et al. (Ex. 1032) as guiding “[t]he
`choice of junctional sites between the mHR and human IgG sequences”;
`Capon’s work is described as “demonstrat[ing] that the joining of the
`molecules near the hinge region resulted in chimeric molecules that were
`both efficiently synthesized and dimerized in the absence of any light chain
`production.” Id. at 2224.
`2. Smith (Ex. 1004)
`Smith teaches DNA sequences encoding human tumor necrosis factor
`receptors (TNF-R), see Ex. 1003, 2:38–41, recombinant expression vectors
`comprising these DNA sequences, and also isolated or purified protein
`compositions comprising soluble forms of TNF-R. Id. at 2:59–61. Smith
`also states that
`A recombinant chimeric antibody molecule may also be
`produced having TNF-R sequences substituted for the variable
`domains of either or both of the immunogl[o]bulin molecule
`heavy and light chains and having unmodified constant region
`domains. For example, chimeric TNF-R/IgG1 may be produced
`from two chimeric genes—a TNF-R/human κ light chain
`chimera (TNF-R/Cκ) and a TNF-R/human γ1 heavy chain
`chimera (TNF-R/Cγ-1). Following transcription and translation
`of the two chimeric genes, the gene products assemble into a
`single chimeric antibody molecule having TNF-R displayed
`bivalently. Such polyvalent forms of TNF-R may have
`enhanced binding affinity for TNF ligand.
`Id. at 10:53–66.
`
`11
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`
`3. Zettlmeissl (Ex. 1005)
`Zettlmeissl reports the development of chimeric antibody-like
`molecules consisting of human CD4 extracellular domains fused to different
`portions of human IgG1 heavy chain constant regions. Ex. 1005, 347,
`Abstract. Five different fusion genes, for expressing the different fusion
`proteins, included a “portion encoding the extracellular domain of CD4 . . .
`and [a] 5-amino-acid linker . . . upstream from the CH1, hinge, or CH2
`exons of the human IgG1 gene, or upstream from the CH1 or CH2 exons of
`the IgM gene.” Id. at 348. Zettlmeissl observed poor expression “for fusion
`proteins bearing CH1 domains.” Id.
`
`E. Obviousness of the Challenged Claims
`Petitioner contends that each of the challenged claims is unpatentable
`as obvious over (1) Watson in view of Smith, and (2) Smith in view of
`Zettlmeissl and Watson. Pet. 1, 9. In the first ground, Petitioner sets forth a
`combination in which the portion of the IgG heavy chain used in Watson is
`fused to the 75-kDa extracellular sequence of the 75-kDa TNFR from Smith.
`Id. at 5 (citing Ex. 1002 ¶¶ 132–144); see also id. at 28–43. In the second
`ground, Smith’s TNFR:IgG fusion protein is modified by deleting the light
`chain and CH1 region of the heavy chain so that only a portion of the IgG
`heavy chain is used in light of the teaching in Zettlmeissl and Watson that
`doing so results in optimum expression. Id. at 5 (citing Ex. 1002 ¶¶ 145–
`161); see also id. at 43–64.
`Petitioner relies on Zettlmeissl and Watson as teaching the use of the
`same, identical portion of the IgG heavy chain as that taught in the ’522
`patent, and relies on that portion for use in the claimed fusion protein. Id. at
`5 (citing Ex. 1002 ¶¶ 76–78, 84–86, 132). As explained below, however, the
`
`12
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`portions of the IgG heavy chain used in Zettlmeissl and Watson—and in
`particular the hinge regions—are not identical to each other or what is taught
`in the ’522 patent. Thus, by asserting that the fusion protein of Watson
`could be modified by adding the 75-kDa extracellular sequence of the TNFR
`from Smith or that Smith could be modified with either the heavy chain
`portion of Zettlmeissl or Watson, Petitioner is unclear what it considers to be
`“all of the hinge . . . domain[],” for either construct under our interpretation
`of this phrase as “all of the hinge, CH2, and CH3 domains.”
`
`1. Obviousness over Watson and Smith
`Petitioner contends that claims 1–10 are unpatentable as obvious over
`Watson in view of Smith. Pet. 28–43. Petitioner further contends that the
`case of obviousness cannot be overcome by objective indicia of
`nonobviousness. Id. at 51–64. Patent Owner opposes Petitioner’s
`contention of obviousness (Prelim. Resp. 31–56) and contends proffered
`objective indicia of nonobviousness confirm the patentability of the
`invention (id. at 65–82).
`Petitioner contends that “Watson’s fusion protein is identical to the
`fusion protein of the ’522 patent claims, except that the receptor protein is
`different.” Pet. 30. Petitioner argues that “[t]he straightforward application
`of Watson’s method to the 75-kDa TNFR disclosed by Smith (i.e., joining
`the extracellular receptor to the hinge-CH2-CH3 region of IgG1) results in
`exactly the same polynucleotides and methods claimed in the ’522 patent.”
`Id. at 31 (citing Ex. 1002 ¶¶ 141–142, 158). Petitioner further offers
`reasoning for the combination stating that “Watson . . . indicates that it has
`optimized the location for attaching a receptor to an IgG to make a fusion
`protein,” and that a person of ordinary skill in the art “would have readily
`
`13
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`applied Watson’s optimized technique of attaching the soluble receptor to
`the hinge-CH2-CH3 portion of an IgG1 . . . to improve on Smith’s
`recommendation to prepare a TNFR:IgG1 fusion protein.” Pet. 38–39
`(citing Ex. 1002 ¶¶ 129, 142, 145, 156–158; Ex. 1003, 2224; Ex. 1004,
`10:53–66).
`The problem with Petitioner’s argument as a whole is that it is unclear
`what Petitioner considers to be the complete hinge. That is, Petitioner is
`inconsistent as to where the boundary lies between the first constant domain
`and the hinge. As background, Dr. Burton sets forth the structure of IgG
`(Ex. 1002 ¶¶ 35–39) that is identified as “[a] schematic depiction of an IgG
`immunoglobulin” (Ex. 1006, 12), reproduced below:
`
`
`Ex. 1002 ¶ 36 (stating the schematic is “[a]dapted from Ex. 1006, p. 12”). In
`defining the structure and function of IgG, Dr. Burton states:
`The hinge region is located between the CH1 and CH2 domains
`of the heavy chain. The hinge region contains all of the
`interchain disulfide bonds that link the heavy chains together. I
`note that in the human IgG1 molecule there is a third disulfide
`bond (shown above) that links the CH1 domain to the constant
`region of the light chain.
`
`
`14
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`Ex. 1002 ¶ 39 (emphasis added).
`Dr. Burton further states that amino acid sequences were known in the
`prior art, citing Ellison.6 Ex. 1002 ¶ 129 (citing Ex. 1050). Patent Owner
`agrees and also relies on Ellison to teach the hinge domain. See Prelim.
`Resp. 15–17, 36. Ellison discloses that the hinge segment is encoded by a
`single exon providing the following amino acid sequence, which includes
`three cysteine (C) residues as opposed to the two set forth in Dr. Burton’s
`schematic of IgG immunoglobulin set forth above:
`E P K S C D K T H T C P P C P
`Ex. 1050, 4072, Fig. 2. Dr. Burton also discusses Capon (Ex. 1032)7 as
`teaching fusion proteins comprising a soluble fragment of a receptor protein
`and portions of an IgG heavy chain. See, e.g., Ex. 1002 ¶¶ 63–68. Like
`Ellison, Capon (Ex. 1032) discloses the hinge region of IgG1 as including
`three cysteine (C) residues (Ex. 1032, 526, Fig. 1) and that “[t]he hinge
`region of each immunoadhesin [Capon’s fusion protein] contains three
`cysteine residues, one normally involved in disulphide bonding to light
`chain, the other two in the intermolecular disulphide bonds between the two
`heavy chains in IgG” (id. at 526).
`Notwithstanding this conflicting evidence and testimony as to what
`constitutes the complete hinge, Dr. Burton asserts that a person of ordinary
`skill in the art would have understood both Watson’s construct and
`Zettlmeissl’s construct, although using different receptors in their respective
`fusion proteins, to achieve “optimal results by employing the identical
`portion of the IgG heavy chain as claimed in the ’522 patent,” thus,
`
`
`6 Ellison et al., 10 NUCLEIC ACIDS RES. 4017–79 (1982) (Ex. 1050)
`7 Capon et al., 337 NATURE 525–31 (1989) (Ex. 1032) (“Capon (Ex. 1032)”).
`
`15
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`containing “all of the domains of the human IgG heavy chain except for the
`first constant region.” Pet. 5 (citing Ex. 1002 ¶¶ 76–78, 84–86, 132); see Ex.
`1002 ¶ 106 (citing Ex. 1003, 2223, Fig. 1), ¶ 105 (citing Ex. 1005, 348, Figs.
`1–2).
`The problem with Petitioner’s argument that is evident from the
`respective disclosures of Zettlmeissl and Watson is that they do not, as
`contended, employ the identical portion of the IgG heavy chain. See Prelim.
`Resp. 24–38 (explaining Watson’s fusion protein does not use the complete
`hinge-CH2-CH3 portion of the IgG1 heavy chain because it lacks the first
`five amino acid sequences of the hinge including the first cysteine residue),
`64 (explaining Zettlmeissl includes an artificial linker between the receptor
`and immunoglobulin sequences). In Zettlmeissl, the sequence “encoding the
`extracellular domain of CD4 . . . and the 5 amino-acid linker sequence were
`placed upstream from the . . . hinge . . . exon[] of the human IgG1 gene.”
`Ex. 1005, 348. The expressed fusion protein, accordingly, would include all
`of the amino acid sequence encoded by the hinge exon of the human IgG1
`gene, but also include a 5 amino-acid linker, which is not identical to
`Watson’s disclosed portion of the IgG heavy chain. Because of the
`transitional phrase “consisting of” that is used to describe the protein
`encoded by the claimed polynucleotide for all claims of the ’522 patent, the
`addition of the 5 amino-acid linker would result in a protein encoded by a
`polynucleotide that is not encompassed by any challenged claim. See AFG
`Indus., Inc. v. Cardinal IG Co., 239 F.3d 1239, 1244–45 (Fed. Cir. 2001)
`(holding the phrase “consisting of” is a “closed” transition phrase that is
`“understood to exclude any elements, steps, or ingredients not specified in
`
`16
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`the claim”). Zettlmeissl does not further define the amino acid sequence of
`the hinge region of IgG1.
`
`Watson’s fusion protein, in contrast, and as contended by Petitioner,
`includes only the two cysteine residues involved in joining the heavy chains,
`i.e., the cysteine residues separated by two proline (P) residues as shown in
`the sequence set forth above as disclosed by Ellison. Pet. 29–33; Ex. 1002
`¶¶ 76–82. The other cysteine, normally involved in intermolecular bonding
`to light chain, is, according to Dr. Burton, part of the CH1 domain. Ex. 1002
`¶ 39 (“[I]n the human IgG1 molecule there is a third disulfide bond . . . that
`links the CH1 domain to the constant region of the light chain.”). As Patent
`Owner points out with regard to the Watson and Smith combination,
`“Petitioner repeatedly emphasizes the supposed criticality of using the same
`IgG portion as the Watson protein to achieve its supposed benefits” to
`achieve the claimed fusion protein. See Prelim. Resp. 38.
`Also, with regard to the Smith, Watson, and Zettlmeissl combination,
`we agree with Patent Owner that Petitioner “incorrectly contends that
`‘modifying Smith’s TNFR:IgG fusion proteins as taught by Zettlmeissl and
`Watson results in the exact fusion proteins recited in the ’522 Patent.”
`Prelim. Resp. 38, 63, respectively; see also Pet. 18 (“Both Watson and
`Zettlmeissl expressly directed the POSA to choose exactly the
`immunoglobulin fragment claimed in the ’522 patent—the ‘hinge-CH2-
`CH3’ region of a human IgG heavy chain.”). We agree with Patent Owner
`that “[f]ollowing Petitioner’s rationale of replacing only the ‘receptor’ in its
`proposed combination of Smith with both Zettlmeissl and Watson would
`thus yield fusion proteins that include a third component (i.e., an artificial
`
`17
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`‘linker’) between the receptor and IgG components [with Zettlmeissl] and
`omit part of the hinge [with Watson].” Prelim. Resp. 64.
`Thus, with respect to Zettlmeissl, Petitioner appears to assert that “all
`of the hinge domain” requires the hinge segment encoded by the hinge exon,
`including three cysteine residues, but with respect to Watson, Petitioner
`appears to assert that “all of the hinge domain” simply requires a portion of
`sequence that includes the two cysteine residues involved in joining the
`heavy chains. Petitioner cannot have it both ways, particularly without an
`explanation why.
`Petitioner fails to provide sufficient explanation why a person of
`ordinary skill in the art reading Watson, which cites to Capon as to the
`sequences and methods used (Ex. 1003, 2222), would understand the hinge
`region of human IgG1 to be any less than that identified by Capon, which
`includes three cysteine residues (Ex. 1032, 526), one of which Dr. Burton
`states is part of the CH1 domain (Ex. 1002 ¶ 39). Watson depicts the
`structure of its mHRLEC fusion protein in Figure 1A and describes what is
`depicted, thusly: “This truncated protein is joined to a human heavy chain
`gamma-1 region immediately NH2-terminal to the hinge domain (H) such
`that this chimera contains the two cysteine residues (C) of the hinge
`responsible for immunoglobulin dimerization as well as the CH2 and CH3
`constant regions.” Ex. 1002, 2223, Fig.1. The cited figure legend further
`identifies other elements included in the figure using other like notation, e.g.,
`(mHR), (SS), (CBD), and (TMD). Id. Thus, the description reasonably
`identifies where the included hinge domain sequence is in the construct, but
`does not convey that the hinge portion included in the construct was
`complete.
`
`18
`
`

`

`IPR2017-01916
`Patent 8,163,522 B1
`Other prior art of record similarly refers to the hinge region as
`including all three cysteine residues. Byrn (Ex. 1033), cited in the Petition
`as prior art “[i]n addition to the prior art relied upon in Coherus’s grounds of
`unpatentability” (Pet. 26), has many authors in common with Capon
`(Ex. 1032), and likewise identifies the hinge region as including three
`cysteine residues (Ex. 1033, 668, Fig. 1). Byrn also teaches that the use of
`the label “Hinge” does not necessarily convey the presence of three cysteine
`residues. For example, Byrn depicts a fusion protein (labeled “CD4
`Immunoadhesin”) with a hinge (labeled “Hinge”) having only the two
`cysteine residues and describes the fusion as joining the CD4 protein
`element to “the first residue in the IgG1 hinge after the cysteine residue
`involved in heavy-light chain bonding.” Id. (emphasis added).
`Watson’s guidance is further insufficient as to what is included from
`“the domains of the constant region of a human immunoglobulin IgG heavy
`chain.” Specifically, even assuming that a person of ordinary skill would
`have understood from Watson that the first domain and hinge domain do not
`correspond to the encoded CH1 region and hinge region, respectively,
`Petitioner has not explained adequately how Watson provides sufficient
`guidance as to the sequence of the disclosed hinge region included to
`provide a fusion protein including all of the heavy chain constant region
`other than the first domain. See generally Pet. Petitioner cites Watson for
`“explain[ing] that ‘[t]he choice of junctional sites between the mHR and
`human IgG sequences was guided by work with human CD4-IgG chimeras
`that demonstrated that the joining