`IN THE UNITED STATES PATENT AND TRADEMARK OFFICE
`_____________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`_____________________
`
`
`MODERNATX, INC.,
`Petitioner
`
`v.
`
`CUREVAC AG,
`Patent Owner
`
`_____________________
`
`Case IPR2017-02194
`Patent 8,383,340
`_____________________
`
`MODERNATX, INC.’S REPLY TO PATENT OWNER’S RESPONSE
`
`
`
`
`
`
`
`
`
`Mail Stop “PATENT BOARD”
`Patent Trial and Appeal Board
`U.S. Patent and Trademark Office
`P.O. Box 1450
`Alexandria, VA 22313-1450
`
`
`

`

`
`
`
`
`
`ModernaTX’s Reply
`IPR2017-02194
`
`TABLE OF CONTENTS
`
`
`I.
`II.
`
`OVERVIEW .................................................................................................... 1
`THE ’340 PATENT WOULD HAVE BEEN OBVIOUS IN VIEW OF THE
`CITED ART ..................................................................................................... 5
`A. CureVac cannot credibly deny that a POSA would have had a reason to
`carry out preparative scale RNA purification ....................................................... 5
`B. Lloyd expressly teaches preparative scale purification of nucleic acids using
`IP RP HPLC with porous non-alkylated PSDVB ................................................. 6
`C. The art did not teach away from the claimed method ................................... 10
`D. A POSA would have had a reasonable expectation of success in combining
`the cited references because DNA and RNA behave similarly .......................... 12
`E. CureVac’s construction of “up to 100 to 10000 nucleotides” reads “up to”
`out of the claims .................................................................................................. 16
`F. The POR fails to show unexpectedly superior results ................................... 16
`III. GJERDE I ANTICIPATES THE CLAIMED METHOD ............................. 18
`A. The intrinsic evidence shows that the proper construction of a “porous
`reversed phase” is a reversed phase characterized by pores. .............................. 19
`B. The preamble is not limiting .......................................................................... 20
`C. Gjerde I explicitly teaches porous substrates ................................................ 22
`D. Gjerde I’s monolith contains pores under either construction of “porous” .. 23
`IV. CUREVAC’s OFFERED DECLARATIONS SHOULD BE GIVEN
`LITTLE WEIGHT ......................................................................................... 24
`CONCLUSION .............................................................................................. 25
`
`V.
`
`
`
`
`- i -
`
`

`

`
`
`I. OVERVIEW
`Moderna’s Petition demonstrated that the ’340 patent’s claims would have
`
`ModernaTX’s Reply
`IPR2017-02194
`
`
`
`been obvious or anticipated by prior art that taught or suggested RNA purification
`
`using porous nonalkylated PSDVB medium in ion-paired reverse phase HPLC (IP
`
`RP HPLC). Nothing in CureVac’s Reply defeats this showing. Though it is
`
`inconvenient for CureVac, the art irrefutably taught purifying nucleic acids using
`
`IP RP HPLC on both preparative and analytical scales, using porous or nonporous
`
`polystyrenedivinylbenzene (PSDVB) as the reversed phase, and using alkylated
`
`and non-alkylated PSDVB.
`
`CureVac alleges that the claimed method “is distinguished [from the prior
`
`art] in that the RNA is purified by means of HPLC using a porous reversed phase
`
`as stationary phase” and that the art “does not teach preparative scale purification.”
`
`EX1001, 1:56-62; POR, pp. 19-30. But CureVac fails to consider what the cited
`
`references as a whole teach, focusing instead only on what has been experimentally
`
`demonstrated. For example, in alleging that Lloyd does not teach preparative scale,
`
`CureVac ignores Lloyd’s explicit disclosure that her IP RP HPLC method is “an
`
`obvious choice” to “produce large quantities” and is “acceptable for…small-scale
`
`preparative work.” EX1005, 227-228. In a similar effort to blunt the obviousness
`
`case against the ’340 patent, CureVac seeks to dismiss Zhang by arguing that it
`
`does not teach RNA purification methods. But a POSA would not have dismissed
`
`
`
`- 1 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`Zhang because Zhang reflects a POSA’s motivations to purify RNA on a
`
`preparative scale, e.g., for use in pre-clinical research and clinical trials.
`
`CureVac’s faulty view of the prior art is further apparent in the Response’s
`
`treatment of DNA and RNA. Dr. Hornby and Dr. Švec agree that the relevant field
`
`is nucleic acid biochemistry, not solely RNA biochemistry. EX1002, ¶¶27-29;
`
`EX2001, ¶32. And a POSA would have recognized the high degree of similarity
`
`between DNA and RNA. EX1070, ¶14. Yet, CureVac pretends that the difference
`
`(-H versus -OH) between these two nucleic acids would preclude a reasonable
`
`expectation of success in using the same purification method for each nucleic acid.
`
`CureVac can reach such a conclusion only by ignoring multiple prior art DNA
`
`purification methods that had also been used to purify RNA as illustrated, e.g., by
`
`Georgopoulos
`
`(EX1023) and Azarani
`
`(EX1008). Even CureVac’s own
`
`employee/witness, Dr. Schwenger, acknowledged that he previously has used the
`
`same HPLC column to purify both DNA and RNA. EX1067, 19:15-19. Rather than
`
`grapple with these inconvenient truths, CureVac relies upon Dr. Schwenger’s 2018
`
`data to purportedly show that one arbitrarily-selected DNA purification column
`
`was not successful in purifying RNA in his hands. Such 2018-generated data
`
`evince nothing about what a POSA would have reasonably expected in 2006, and
`
`simply cannot refute prior art evidence showing that various purification methods,
`
`including IP RP HPLC methods using PSDVB as the reversed phase, had been
`
`
`
`- 2 -
`
`

`

`
`
`
`successfully used to purify both DNA and RNA.
`
`ModernaTX’s Reply
`IPR2017-02194
`
`Instead of accepting that the art was well-advanced by 2006, CureVac feigns
`
`inability on the part of POSAs to overcome routine challenges, even for
`
`rudimentary scientific techniques such as mitigating nuclease activity while
`
`extracting and isolating RNA. POR, pp. 41-44. But such techniques were so well
`
`established by 2006 that they were included in basic laboratory manuals, such as
`
`Sambrook’s iconic “Molecular Cloning A Laboratory Manual.” EX1054; EX1071;
`
`EX1070, ¶¶28-31. Again, CureVac has failed to view the art through a POSA’s
`
`eyes.
`
`Similarly, CureVac fails to apply a POSA’s perspective in construing
`
`claims. CureVac construes the term “porous reverse phase” to improperly import
`
`the limitation of requiring pore sizes that allow RNA molecules to enter. But a
`
`POSA would have understood the intrinsic evidence to show that “porous” simply
`
`means “characterized by pores.” EX1001, 7:1-3; see §III.A below. CureVac used
`
`the broad explicit definition of “porous” when it sought broad patent protection,
`
`but now, when faced with damaging prior art, CureVac belatedly seeks to impose
`
`additional limitations on the claims. CureVac also now tries to avoid prior art by
`
`limiting the claims’ preamble, arguing that the method is limited to preparative
`
`scale purification. But the preamble merely states an intended use for the claimed
`
`method. See §III.B below. And no step or limitation in the claim body restricts the
`
`
`
`- 3 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`claim to a preparative scale method. The present case thus presents a classical
`
`example of a non-limiting preamble used “only to state a purpose or intended use
`
`for the invention." Georgetown Rail Equip. v. Holland L.P., 867 F. 3d 1229, 1236
`
`(Fed. Cir. 2017). Absent its faulty claim construction arguments, CureVac’s
`
`anticipation defense falls apart. See §III below.
`
`To the extent that CureVac attempts to establish objective indicia of non-
`
`obviousness, those attempts do not comport with the proper legal standards. See
`
`§II.F below; EX1070, ¶¶108-128. CureVac’s alleged unexpectedly superior results
`
`evidence does not even purport to compare CureVac’s method with the closest
`
`prior art, let alone show that CureVac’s results were unexpected. Rather, CureVac
`
`compared RNA produced by its method with RNA that was not purified or RNA
`
`that was not HPLC-purified—even though HPLC-based RNA purification methods
`
`were known in the art. It is not surprising that CureVac claimed superiority when it
`
`set the bar so low. And even if, for argument’s sake, one were to accept that the
`
`results obtained were superior, CureVac has failed to show a nexus with the
`
`claimed invention, and failed to show that such results are commensurate in scope
`
`with the claimed invention. In a similar vein, CureVac has not shown that Dr.
`
`Fotin-Mleczek’s testimony regarding CureVac’s place in the history of RNA-based
`
`therapeutics relates to any legal principle relevant to obviousness. CureVac’s
`
`arguments fail.
`
`
`
`- 4 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`II. THE ’340 PATENT WOULD HAVE BEEN OBVIOUS IN VIEW OF
`THE CITED ART
`
`CureVac’s responsive arguments do not distinguish between Grounds 2-5.
`
`And, at bottom, the arguments are legally and factually flawed. The Board
`
`therefore should find that all claims of the ’340 patent would have been obvious.
`
`EX1070, ¶¶8-10, 50-128.
`
`A. CureVac cannot credibly deny that a POSA would have had a reason to
`carry out preparative scale RNA purification
`
`Before 2006, the art recognized a need to purify large amounts of RNA for
`
`preclinical use and for clinical trials. EX1070, ¶¶8, 20, 51-54. The art disclosed
`
`testing RNA-based agents for clinical use before 2006, and Zhang and Sullenger
`
`discuss some of those RNA therapeutics, including ribozymes, siRNA, and
`
`mRNA. EX1038, 11, 22; EX1039; EX1070, ¶¶8, 15-20, 51-54, 102; EX1059, 1;
`
`EX1060, 405; EX1061, 1635, 1636, 1638-1639. As explained in the Petition, a
`
`POSA would have known that “kilograms [of RNA] are required” for clinical
`
`development of RNA-based therapeutics. Petition, pp. 42-43; EX1047, 203-204;
`
`EX1051, 94:14-21; EX1049, ¶¶[204], [354]-[355]. Rather than considering the
`
`need for RNA-based therapeutics as taught in the prior art, CureVac improperly
`
`analyzes Zhang’s “teaching alone” and dismisses it because it doesn’t discuss
`
`purification. POR, pp. 32-33. CureVac’s piecemeal approach is improper and
`
`inconsistent with the art’s actual comprehensive teachings.
`
`
`
`- 5 -
`
`

`

`
`
`
`ModernaTX’s Reply
`IPR2017-02194
`
`CureVac also mischaracterizes the Petition’s arguments by presenting a
`
`strawman argument that the Petition’s obviousness arguments are linked only to
`
`Zhang’s teachings relating to “AS-ODN.” POR, pp. 35-37. But the Petition pointed
`
`to Zhang’s teachings about the developing “antisense therapeutics” field, in
`
`reference to the multiple RNA-based therapeutics described in Zhang (or in
`
`Sullenger). EX1038, 11; Petition, pp. 42-44, 53-55; EX1070, ¶¶51-54, 102-103.
`
`The Petition’s arguments were not focused on Zhang’s AS-ODN discussion, and
`
`any reference to DNA-based therapeutics does not detract from Zhang’s teachings
`
`about therapeutic uses of RNA. EX1070, ¶¶51-54.
`
`CureVac also incorrectly alleges that Moderna ignores “Gjerde I’s
`
`suggestion to use a non-porous reverse phase.” POR, pp. 34-35. This argument
`
`holds no merit because it is illogical to suggest that Moderna ignored the reference
`
`asserted in the Petition’s anticipation ground. Further, the presence of an
`
`alternative method in the art does not detract from Lloyd’s teachings, especially
`
`when Lloyd published several years later and presented its porous-media based
`
`method as the “obvious choice” because of the advantages it offered. EX1005,
`
`227; EX1070, ¶¶69-73.
`
`B. Lloyd expressly teaches preparative scale purification of nucleic acids
`using IP RP HPLC with porous non-alkylated PSDVB
`
`CureVac attempts to rebut the obviousness grounds by narrowly arguing that
`
`Lloyd, Zhang, and Sullenger do not perform preparative scale RNA purification.
`
`
`
`- 6 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`POR, pp. 1, 25-32, 57. But CureVac disregards Lloyd’s express recommendations
`
`to use her method for preparative-scale oligonucleotide purification. Petition, pp.
`
`41-42. Lloyd states that her method is "an obvious choice" to "produce large
`
`quantities of well-defined oligonucleotides… [because it has] excellent selectivity
`
`and high capacity." EX1005, 227. And Lloyd further notes that its IP RP HPLC
`
`method is “acceptable for…small-scale preparative work.” EX1005, 228. CureVac
`
`characterizes Lloyd’s conclusions and recommendations regarding large-scale
`
`purification as a “desire.” POR, p. 26. But Lloyd need not carry out a
`
`recommended method to render it obvious.
`
`Lloyd’s experiments and data provide
`
`support
`
`for
`
`its express
`
`recommendations
`
`to perform preparative purification. CureVac
`
`improperly
`
`dismisses these experiments as irrelevant by focusing on experimental details
`
`without addressing Lloyd’s explanations of why those experiments are significant.
`
`EX1070, ¶¶55-68. For example, CureVac dismisses Lloyd’s loading capacity test
`
`as irrelevant because is “not a teaching of preparative HPLC, or separation of any
`
`kind.” POR, p. 27. But capacity is highly relevant: as Lloyd states, it is “necessary
`
`to know” column capacity because “[f]or purification it is essential…to have high
`
`capacity [to] enable[e] rapid, economic purifications to be achieved.” EX1005, 225
`
`(emphasis added); EX1070, ¶¶64-68. CureVac presents this narrow focus, despite
`
`arguing that “low loading capacity [columns are] only suitable for analytical
`
`
`
`- 7 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`separations.” POR, p.4. A second example of CureVac’s improperly dismissing
`
`Lloyd’s data is CureVac’s insistence that Lloyd’s IP RP HPLC oligonucleotide
`
`separation is “[n]ot [p]reparative [p]urification.” POR, pp. 28-30; EX1005, Figure
`
`1. But Lloyd’s experiments are highly relevant because “[f]or purification it is
`
`essential to achieve resolution,” and, as Dr. Švec admits, Lloyd’s experiments “tell
`
`a POSA what molecules may be separated when using the column.” EX2016, ¶27;
`
`EX1070, ¶¶57-63. EX1005, 225-228.
`
`CureVac also attempts to minimize Lloyd’s teachings by narrowly focusing
`
`on whether Lloyd expressly teaches purifying RNA. POR, pp. 25-26. But a POSA
`
`would have known that Lloyd teaches using RNA because Lloyd describes its
`
`results using the term “oligonucleotides,” a broad term that encompasses both
`
`DNA and RNA. See, e.g., EX1005, Abstract, 227; EX1070, ¶¶8, 53; EX1002,
`
`¶158; EX1005, 227:1:2; EX1064, 90; EX1024, 101, 105. “Since all of the nucleic
`
`acids may be regarded as polymers of nucleotides, they are often referred to by the
`
`generic name polynucleotides… [or] oligonucleotides.” EX1064, 90 (emphasis
`
`added); EX1035, G-17; EX1070, ¶53. Dr. Švec uses the term oligonucleotide
`
`generically in his own work, such as in his review of “[h]igh-resolution separations
`
`of oligonucleotides” in which he discusses both DNA and RNA separations.
`
`EX1073, §3.2. Lloyd’s intent to use the term oligonucleotide generically is
`
`confirmed by Lloyd’s citation to RNA-related references when discussing
`
`
`
`- 8 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`oligonucleotides. EX1001, ¶158; EX1005, 224, 228. Lloyd’s intent is further
`
`confirmed by her reference to preparing oligonucleotides for “antisense therapies,”
`
`a term commonly associated with RNA. EX1005, 223; EX1070, ¶¶50-53; EX1038.
`
`For example, Zhang discusses a variety of RNA’s as antisense therapies, including
`
`ribozymes and siRNA. EX1038, 11; EX1070, ¶¶51-54. Accordingly, a POSA
`
`would have known that Lloyd’s teachings were applicable to both DNA and RNA.
`
`EX1070, ¶53.
`
`Contrary to CureVac’s suggestion that Lloyd’s ion-exchange experiments
`
`are “inapposite,” the quotes highlighted by CureVac demonstrate that both ion
`
`exchange and IP RP HPLC were possible methods for preparative purification.
`
`POR, pp. 30-32; EX1070, ¶¶69-73. Lloyd states that “it is clear that ion-pair
`
`reversed-phase HPLC can be used for
`
`the analysis and purification of
`
`oligonucleotides.” EX1005 at 228. Lloyd further states that ion exchange is an
`
`“alternative purification strategy” that can be used for preparative purifications if
`
`using organic solvents is not desired. Id. Such a statement of an alternative carries
`
`the implicit acknowledgement that the original will also function for the same
`
`purpose. EX1070, ¶72.
`
`CureVac also places much too high a standard on Lloyd’s data by describing
`
`an ideal chromatogram and requiring that Lloyd meet the stringent criteria for that
`
`ideal. POR, pp. 48-53. But Lloyd shows purification as defined in the ’340 patent:
`
`
`
`- 9 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`Lloyd’s data showed that “the desired RNA in a sample is separated and/or isolated
`
`from the impurities present therein.” EX1001, 2:61-64; EX1070, ¶¶37-38, 57-63.
`
`Lloyd successfully resolved at least six different nucleic acid species, and as noted
`
`by Dr. Švec, such a result “ha[s] achieved separation.” EX1058, 45:9-46:14;
`
`EX1070, ¶¶57-63.
`
`C. The art did not teach away from the claimed method
`CureVac presents two teaching away arguments that both incorrectly
`
`characterize an acknowledgement of alternative methods in the prior art as a
`
`teaching away. EX1070, ¶¶69-73.
`
`First, CureVac illogically argues that Lloyd teaches away from using IP RP
`
`HPLC for preparative scale purification, despite Lloyd’s teachings that this method
`
`was “acceptable for…small-scale preparative work” and the “obvious choice” for
`
`preparative purifications. POR, pp. 38-39; EX1005, 227-228. CureVac’s only
`
`justification for asserting that Lloyd teaches away from the technique it expressly
`
`recommends is Lloyd’s acknowledgement that another technique could be used as
`
`an “alternative.” POR, p. 38-39. But Lloyd does not express any doubt as to the
`
`efficacy or usefulness of its IP RP HPLC methods. EX1070, ¶70. To the contrary,
`
`Lloyd states that “it is clear that ion-pair reversed-phase HPLC can be used for the
`
`analysis and purification of oligonucleotides.” EX1005, 228. CureVac’s argument
`
`fails, because “the prior art’s mere disclosure of more than one alternative does not
`
`
`
`- 10 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`constitute a teaching away from any of these alternatives because such disclosure
`
`does not criticize, discredit, or otherwise discourage the solution claimed….” In re
`
`Fulton, 391 F.3d 1195, 1201, (Fed. Cir. 2004).
`
`Second, CureVac mischaracterizes the art in arguing that the art taught away
`
`from using porous media. POR, pp. 39-41. In support, CureVac primarily
`
`references Gjerde I’s preference for nonporous PSDVB media in analytical
`
`purifications. Id. But Lloyd notes that nonporous media function poorly in
`
`preparative purifications because “the capacity, due to the loss of the internal pore
`
`volume, is low and the operating pressures high.” EX1005, 224; EX1070, ¶¶65,
`
`72; EX1065, 29. This is why, as Dr. Švec stated in 2002, a POSA would have
`
`understood that “nonporous supports are preferred only for high speed separation
`
`of peptides, proteins, and nucleic acids on an analytical scale.” EX1065, 29
`
`(emphasis original). Lloyd recommends porous media as an “alternative approach”
`
`that solves those problems because of porous media’s “improved” properties.
`
`EX1005, 224. Therefore, a POSA would have understood that both porous and
`
`nonporous media were options for IP RP HPLC, and that porous media offered
`
`advantages in preparative purifications. EX1070, ¶¶69-73.
`
`CureVac incorrectly alleges that Dr. Hornby’s publications support their
`
`view of porous media. POR, pp. 40-41; EX1070, ¶¶74-76. But neither publication
`
`CureVac references (EX2024 and EX2025) would have informed a POSA before
`
`
`
`- 11 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`the 2006 priority date, because they were published in 2009. Further, Dr. Hornby’s
`
`papers are consistent with the Petition’s arguments because they merely recognize
`
`a fact that neither party disputes: nonporous media had been used for separating
`
`RNA. EX1070, ¶¶75-76.
`
`D. A POSA would have had a reasonable expectation of success in combining
`the cited references because DNA and RNA behave similarly
`
`CureVac’s expectation of success arguments are misleading because they
`
`exaggerate the differences between DNA and RNA, without acknowledging the
`
`extensive evidence that DNA and RNA behave similarly in purification methods.
`
`EX1070, ¶¶81-84. To the extent that CureVac and Dr. Švec provide evidence that
`
`DNA and RNA are “[b]ehaviorally [d]ifferent,” they nonetheless fail to establish
`
`that methods to purify DNA would not work with RNA. Id. There is only a single
`
`atom’s difference in the DNA and RNA backbones. POR, p. 13 (citing EX2027);
`
`EX1071, 847:2:1; EX1058, 39:7-20, 40:5-6. And because of the many structural
`
`similarities between DNA and RNA, the field expected that “[s]eparation
`
`characteristics of RNA are expected to resemble those of ssDNA under fully
`
`denaturing conditions.” EX1008, 1:2:1; EX1023, 110:2:2; EX1004, 27:17-19;
`
`EX2016, ¶12; EX1058, 40:8-11; EX1062, 3:1:1. CureVac is inappropriately
`
`holding the Petition to a standard higher than a reasonable expectation of success
`
`by requiring evidence that “a POSA would have generally understood DNA to be
`
`interchangeable with RNA.” POR, p. 43.
`
`
`
`- 12 -
`
`

`

`
`
`
`ModernaTX’s Reply
`IPR2017-02194
`
`CureVac’s expectation of success arguments are further flawed because they
`
`ignore prior successes in adopting methods to purify DNA for RNA. EX1070,
`
`¶¶21-27, 81-82. For example, in 2000, Georgopoulos applied IP RP HPLC with a
`
`PSDVB reversed phase “to RNA molecules comparable in size to the DNA
`
`molecules separable by this method.” EX1023,110. Similarly, Azarani “extended”
`
`methods to purify DNA with IP RP HPLC using a PSDVB reversed phase “to
`
`RNA analysis.” EX1008, 1. And Gelhaus provided further evidence of IP RP
`
`HPLC’s general applicability for RNA purification, teaching that its method—also
`
`using PSDVB—can differentiate stem-loop conformations of two RNA hairpins
`
`and that RNA’s inclusion of uracil has “minimal” effect. EX1062, Abstract, 3.
`
`Further expectation of success would have been provided by prior art
`
`references teaching basic features of performing IP RP HPLC to purify RNA (e.g.,
`
`the counterion, equipment, reagents, and environmental conditions). EX1070,
`
`¶¶77-80. The Polymer Labs 2004-2005 catalog expressly recommended using its
`
`PLRP-S media for purifying DNA and RNA oligonucleotides. EX1024, 101. And
`
`more generally, a POSA would have known that PLRP-S media provides "easy
`
`scale-up" from analytical to preparative scale. EX1070, ¶¶78-79; EX1002, ¶164;
`
`EX1024, 101; EX 1066, 87; EX1062, 5.
`
`CureVac attempts to blunt the significance of the previously-successful
`
`adoptions of DNA purification methods for RNA by parsing whether individual
`
`
`
`- 13 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`methods differed from the claimed methods by performing an analytical separation
`
`or using nonporous beads. POR, pp. 44-48. But prior art methods need not be
`
`identical to the claimed methods to provide an indication of what is expected to be
`
`successful in the future. And the art as a whole shows a consistent successful
`
`application of DNA purification techniques to RNA. EX1070, ¶¶21-27, 81-82.
`
`In contrast, CureVac has not presented a single example of a prior art DNA
`
`purification method that could not be adapted for purifying RNA. CureVac only
`
`discusses Dr. Schwenger’s assertion that, in his hands, IP RP HPLC with a silica-
`
`based C18-alkylated column was less effective in purifying RNA than DNA.
`
`EX2014, ¶¶45-47. Dr. Schwenger’s experiments are not evidence of a lack of
`
`expectation of success for at least three reasons. EX1070, ¶¶92-96. First, his data
`
`would not have been available to a POSA and thus would not have impacted a
`
`POSA’s expectation of success as of Dec. 22, 2006. Id. Second, Dr. Schwenger’s
`
`results with a single purification method (even if known) would not have
`
`undermined a POSA’s reasonable expectation of success because a single counter-
`
`example would not have been sufficient to undermine a reasonable expectation of
`
`success from multiple successful examples (as discussed above). Id. And third, Dr.
`
`Schwenger’s conclusion that “a silica-based RP matrix…was not suitable for
`
`separating RNA molecules” is belied by Dr. Schwenger’s cross-examination
`
`testimony that he had successfully used a column with a C18-alkylated silica base
`
`
`
`- 14 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`for IP RP HPLC purification of both DNA and RNA while obtaining his doctorate.
`
`EX1067, 13:11-14:3, 15:22-16:14, 17:10-19:19; EX1070, ¶¶92-96.
`
`CureVac’s arguments are further flawed because they do not account for a
`
`POSA only needing a reasonable expectation of achieving some purification and
`
`resolution with Lloyd’s PLPR-S columns. EX1070, ¶¶85-87. The claims do not
`
`require any particular purity level or any particular degree of separation. Id. A
`
`POSA would have reasonably been expected to achieve such success. EX1070,
`
`¶¶85-87. Even CureVac does not require 100% purity.
`
`
`
`
`
`
`
`
`
`EX1068, 90:18-91:15; EX2013, Appendix I.
`
`And contrary to CureVac’s assertions, Dr. Hornby’s past statements do not
`
`undermine his reasonable expectation of success arguments. POR, pp. 41-44;
`
`EX1070, ¶¶88-91. Dr. Hornby’s statements that dealing with RNA was “difficult”
`
`relate to challenges associated with extracting or isolating RNA that had long been
`
`conquered by 2006. EX1070, ¶89; EX1058; EX1072; EX1054, Chapter 7;
`
`EX1058, 64:1-5; EX1072, 7.6-7.9. Similarly, Dr. Hornby’s recognition of
`
`differences in RNA and DNA behavior is immaterial in light of the repeated
`
`demonstrations that the same purifications methods could be applied to both DNA
`
`
`
`- 15 -
`
`

`

`
`
`and RNA. EX1070, ¶91.
`
`
`
`ModernaTX’s Reply
`IPR2017-02194
`
`E. CureVac’s construction of “up to 100 to 10000 nucleotides” reads “up to”
`out of the claims
`
`Moderna disagrees with CureVac’s construction of the term “up to 100 to
`
`10000 nucleotides,” because CureVac proposes a construction that would eliminate
`
`any meaning for the term “up to.” POR, p. 20; EX1070, ¶¶44-49. An RNA “up to”
`
`a length of 100 to 10000 nucleotides means that the claim covers purifying RNAs
`
`with lengths up to a maximum of 100 to 10000 nucleotides, with no limitation on
`
`the minimum length of RNA. EX1070, ¶¶44-49. Support for this interpretation is
`
`seen in the specification, which describes purifying RNA as small as 20
`
`nucleotides. Id.; EX1001, 4:54-67. Dr. Švec agrees that “very short chains” are
`
`included in the meaning of “up to.” EX1058, 83:12-15. Regardless, construing this
`
`term is not necessary to resolve the issues in this case, as the cited references teach
`
`purifying RNA sizes covered by either meaning of claim 4. Petition, pp. 25, 45.
`
`F. The POR fails to show unexpectedly superior results
`CureVac argues that the claimed method produces unexpectedly superior
`
`results, but CureVac’s arguments do not meet the standard for that or any other
`
`objective indicia of nonobviousness. EX1070, ¶¶10, 108-128.
`
`First, CureVac does not even attempt to meet the standard of “show[ing] [the
`
`results] to be unexpected compared with the closest prior art.” EX1070, ¶¶112-113,
`
`123; Kao Corp. v. Unilever United States, Inc., 441 F.3d 963, 970 (Fed. Cir. 2006).
`
`
`
`- 16 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`Dr. Thran and Dr. Fotin-Mleczek present data comparing “HPLC-purified mRNA”
`
`to control mRNA labelled as either “non-purified mRNA” or “non-HPLC purified
`
`mRNA.” EX2013, ¶¶14, 17, 20; EX2012, ¶¶24-26, 31, 35-36. But Dr. Thran and
`
`Dr. Fotin-Mleczek do not describe what, if any, purification or treatment steps
`
`were taken with the control mRNA samples. EX1070, ¶122. CureVac has not
`
`demonstrated that the claimed method is superior to any of the many purification
`
`methods taught in the prior art. EX1070, ¶¶112-113, 123. For example, both
`
`Gjerde I and Lloyd teach purification methods that are closer prior art than “non-
`
`purified” RNA, yet Dr. Thran and Dr. Fotin-Mleczek never even read Gjerde I or
`
`Lloyd, much less compared their methods to the ’340 patent claims. Id.; EX1068,
`
`50:13-51:2; EX1069, 72:1-21.
`
`Second, CureVac’s arguments are inadequate because they have not
`
`established a nexus between the results and the claimed methods. CureVac asserts
`
`that transfecting RNA purified with the ’340 patent’s method results in greater
`
`expression and decreased induction of IL-6, and cites Kariko (EX2031) as
`
`explaining the phenomenon. But Kariko did not use the purification method of the
`
`’340 patent. Rather, Kariko purified mRNA using an alkylated non-porous PSDVB
`
`as the reversed phase. EX2031, 2:1:5-2:1. Thus, Kariko demonstrates that a
`
`purification method similar to those disclosed in the prior art (e.g., in Gjerde I) had
`
`the same effects as the claimed method, and that the alleged superior results are not
`
`
`
`- 17 -
`
`

`

`
`
`
`linked to the ’340 patent’s method. EX1070, ¶¶118, 124.
`
`ModernaTX’s Reply
`IPR2017-02194
`
`And third, CureVac has not met the requirement that “[e]vidence of
`
`secondary considerations must be reasonably commensurate with the scope of the
`
`claims.” In re Huai-Hung Kao, 639 F. 3d 1057, 1068 (Fed. Cir. 2011). CureVac’s
`
`test data relate only to testing mRNA, but the claims cover a wide range of RNA
`
`types that CureVac has not tested such as tRNA, rRNA, siRNA, ribozymes, or
`
`whole-cell RNA. EX1001, 3:64-4:6, Claim 2; EX1070, ¶¶114-117; EX1068, 33:3-
`
`8; EX1058, 27:2-30:18. Further, some RNA types cannot possibly have the alleged
`
`superior property because they are non-coding RNA, and by definition non-coding
`
`RNAs have no expression to be improved. EX1070, ¶¶15-16; EX1054, 855;
`
`EX1055, 231, 263; EX1056, 1260; EX1058, 31-32; EX1068, 36; 38.
`
`Accordingly, the Board should find all claims of the ’340 patent obvious.
`
`III. GJERDE I ANTICIPATES THE CLAIMED METHOD
`With respect to Ground 1, the parties only dispute the construction of
`
`“porous reversed phase” and the preamble’s statement “for purifying RNA on a
`
`preparative scale.” POR, pp. 21-25. But Gjerde I teaches porous reversed phase
`
`regardless of which construction of porous is used, and the claims do not require
`
`preparative scale purification. The Board should find that Gjerde I anticipates
`
`claims 1-5, 8, 10-22, and 26 of the ’340 patent. EX1070, ¶¶11, 129-137.
`
`
`
`- 18 -
`
`

`

`
`
`
`ModernaTX’s Reply
`IPR2017-02194
`
`A. The intrinsic evidence shows that the proper construction of a “porous
`reversed phase” is a reversed phase characterized by pores.
`
`Moderna respectfully submits
`
`that
`
`the Board should reconsider
`
`its
`
`preliminary construction of “porous reversed phase” as a reversed phase that is
`
`“‘characterized by pores’ that allow the RNA ‘molecules to get inside the pores of
`
`the matrix.’” Institution Decision, Paper 9, pp. 8-10.
`
`The specification explicitly defines
`
`the
`
`term porous, stating
`
`that,
`
`“[i]rrespective of its precise nature, the polymeric stationary phase is porous in its
`
`nature, which means that the beads or the block are characterized by pores.”
`
`EX1001, 7:1-3 (emphasis added). A patentee can act as a lexicographer, and has
`
`done so here. Accordingly, the broadest reasonable construction of “porous
`
`reversed phase” is a reversed phase characterized by pores. EX1070, ¶¶39-43. The
`
`specification’s explicit definition does not reference, much less require, a particular
`
`size or type of pore. Id. Including limitations linking pore size and RNA size
`
`amounts to importing a limitation from the specification.
`
`CureVac’s flawed construction of porous depends on the specification’s
`
`reference to “a possibility” that RNA entering the pores “provide[s] a further
`
`retention effect.” But this falls far short of requiring that the pores allow the RNA
`
`molecules to get inside the pores, which would mean that the required pore size
`
`depends on the size of the RNA to be purified. EX1001, 7:50-54; EX1070, ¶39.
`
`While the ’340 patent notes that “the pore size of the reversed phase may also be
`
`
`
`- 19 -
`
`

`

`ModernaTX’s Reply
`
`IPR2017-02194
`
`
`selected in dependence of the size of the RNA to be separated,” it does not state
`
`that the pore size must be selected based on the RNA to be separated. EX1001,
`
`7:45-54 (emphasis added); EX1070, ¶42. Nor does the