Trials@uspto.gov
`571-272-7822
`
`
`Paper No. 56
`
`
`UNITED STATES PATENT AND TRADEMARK OFFICE
`____________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`____________
`
`PFENEX, INC.,
`Petitioner,
`
`v.
`
`GLAXOSMITHKLINE BIOLOGICALS SA,
`Patent Owner.
`____________
`
`IPR2019-01028
`Patent 9,422,345 B2
`____________
`
`Record of Oral Hearing
`Held: September 9, 2020
`____________
`
`Before SHERIDAN K. SNEDDEN, JO-ANNE M. KOKOSKI, and
`RICHARD J. SMITH, Administrative Patent Judges.
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`APPEARANCES:
`
`ON BEHALF OF THE PETITIONER:
`
`
`JEFFREY GUISE, ESQUIRE
`LORELEI P. WESTIN, ESQUIRE
`ALICIA LIPOSHIK, ESQUIRE
`Wilson Sonsini Goodrich & Rosati
`12235 El Camino Road
`San Diego, California 92130-3002
`
`LORA M. GREEN, ESQUIRE
`Wilson Sonsini Goodrich & Rosati
`1700 K Street, NW.
`Washington, D.C. 20006
`
`WENDY L. DEVINE, ESQUIRE
`Wilson Sonsini Goodrich & Rosati
`One Market Plaza
`Spear Tower, Suite 3300
`San Francisco, California 94105
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`2
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`ON BEHALF OF THE PATENT OWNER:
`
`
`CHARLES E. LIPSEY, ESQUIRE
`Finnegan, Henderson, Farabow, Garrett, & Dunner, LLP
`Two Freedom Square
`11955 Freedom Drive
`Reston, Virginia 20190-5675
`
`AMANDA K. MURPHY, ESQUIRE
`TRENTON A. WARD, ESQUIRE
`YIEYIE YANG, ESQUIRE
`Finnegan, Henderson, Farabow, Garrett, & Dunner, LLP
`901 New York Avenue, NW.
`Washington, D.C. 20001-4413
`
`
`The above-entitled matter came on for hearing on Wednesday,
`September 9, 2020, commencing at 1:00 p.m. EDT, by video/by telephone.
`
`
`
`
`3
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`
`P R O C E E D I N G S
`- - - - -
`
`JUDGE SMITH: Good afternoon this is the oral hearing in IPR2019-
`01028. I am Judge Smith and I am joined by Judge Snedden and Judge
`Kokoski. May we have appearances for the parties beginning with the
`petitioner.
`MR. GUISE: Morning your honors, this is Jeff Guise for petitioner
`Pfenex. We have several people in the room that you cannot see. We have
`Lori Westin, we have Alicia Liposhik, we have Lora Green on the phone.
`And we also have, from the company, we have Diane Retallick who is on
`the audio line. Today, Wendy Devine is going to do the argument for the
`petitioner and thank you very much for listening to us.
`MS. DEVINE: Good Morning
`JUDGE SMITH: Good Morning. Patent Owner.
`MS. MURPHY: Good afternoon your honors, I am Amanda Murphy
`appearing on behalf of patent owner Glaxosmithkline. I will be arguing
`patent education sheet on co construction and non obviousness of the
`original endomomic claim. We would like to thank your honors for
`allowing my colleague Dr. Yang to join me today under the office of LEAP
`program. Dr. Yang will be arguing the statutory requirement of patent
`owner’s motion to amend. I have in the room with me today Charles Lipsey
`and Trenton Ward. We understand that in view of Dr. Yang’s participation
`we will receive an additional 15 minutes of arguing time. We will reserve
`15 minutes for our rebuttal.
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`
`
`
`4
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`
`JUDGE SMITH: Thank you. We set forth a procedure in how we are
`going to handle the oral hearing today in our order dated August 5th. As a
`reminder we are going to go over a few things before we get started. First, a
`public audio link has been requested for this hearing and I just learned that
`apparently it is active. The parties have not identified any confidential
`information to be discussed during the hearing but please be aware that other
`persons may be listening to the hearing. Petitioner has the burden of
`showing unpatentability of the challenged claims and will go first. Patent
`Owner will then have the opportunity to respond. Each party has requested
`and been granted 60 minutes total time with the exception, as noted, that
`patent owner has requested and been granted an additional 15 minutes to
`accommodate Ms. Yang as the LEAP attorney. A party may reserve some
`of its time for rebuttal. Please let us know before you begin how much
`rebuttal time you would like and when you would like a warning that your
`time is almost up. Judge Snedden will be monitoring our time today.
`A couple of pointers regarding speaking. Be sure to unmute yourself,
`but only when you are speaking. Identify yourself each time you speak.
`Observe a pause prior to speaking because there may be a delay in
`transmission of words spoken by the previous speaker. This is intended to
`avoid speaking over others. Identify clearly and specifically, each
`demonstrative or exhibit you refer to, such as by slide number or exhibit
`number and page, so that the court reporter’s transcript will be clear and
`accurate. And to permit the judges to follow the arguments. Petitioner
`would you like to reserve some time?
`MS. DEVINE: Thank you your honor. We would like to reserve 20
`minutes please.
`
`
`
`5
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`
`JUDGE SMITH: And what type of warning would you like?
`MS. DEVINE: 15 minutes.
`JUDGE SMITH: Okay you may begin when ready.
`MS. DEVINE: Thank you. May it please the court. My name is
`Wendy Devine and I represent the petitioner Pfenex. Turning to slide 2 of
`our demonstrative, the present matter involves the unpatentability of a
`certain claim of US Patent 9,422,345 for which the earliest possible priority
`date is in October 2009.
`Now turning to slide 4. Here we presented the 2 independent claims
`of the 345 patent. These claims cover the polynucleotide. Now importantly
`these claims do not cover the method. The claimed polynucleotide broadly
`requires 3 prime portion and 5 prime signal. That 3 prime portion it
`becomes a mature bacterial toxin polypeptide having a sequence that is at
`least 90 percent identical to a mutant diphtheria toxin commonly referred to
`as CRM197. The second piece the 5 prime signal sequence is a sequence
`that is not derived from diphtheria that encodes a polypeptide having a
`sequence that when expressed in a bacterial host cell, is capable of directing
`transport of the first piece, the bacterial toxin polypeptide to the periplasm of
`the host.
`Now turning to slide 53. Here we have presented the first proposed
`amended claim. The proposed amended claims do not change the structure
`of the polypeptide. At most proposed amended claims as functional
`characteristics of the 5 prime signal sequence including the direct transport
`of the CRM197 or CRM197 like protein. The specified amount of the
`protein expression is in the periplasm and soluble. And that standard
`techniques such as ELISA can be used to detect the expressed protein. Now
`6
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`the only potential structural change accomplished by the proposed
`amendment is the change from at least 90 percent CRM197 to consisting
`essentially of CRM197. Now its not clear from the record what the full
`scope of consisting essentially of CRM197 is, but regardless patent owner
`had not explained how this amendment differentiates the prior art. And in
`fact, this does not differentiate the prior art because the prior art including
`Zhou and Davis teach CRM197 itself. So, let the record reflect that the 345
`patent is not the first disclosure of the periplasmic expression of
`heterologous protein. Let the record also reflect that the 345 patent is not the
`first disclosure of signal sequences that function to express soluble protein in
`the periplasm of the host bacteria. And the record also reflect that the 345
`patent is not the first disclosure of CRM197. Now notably, CRM197 was
`developed all the way back in 1971. And differs from Wild Type diphtheria
`toxin only with one active substitution in at position 52. Now CRM197 like
`other diphtheria toxins is a disulfide linked protein. Once expressed it needs
`to disulfide bond to fold correctly. Now this is important because CRM197
`has been the subject of a lot of research because of its properties. And those
`properties include that it doesn’t have the toxicity of Wild Type Diphtheria
`toxin. But it does invoke the immune response to Wild Type Diphtheria
`toxin. And in order to accomplish those things it has to fold correctly.
`So, moving on. Against this backdrop, if we turn to slide 13. The
`prior art motivates the expression of CRM197 in the periplasm of host
`bacteria. In slide 13 we excerpt exhibit 1029. And 1029 is a 1987 paper that
`discusses many advantages of secretion in E. coli known back in the 1987
`timeframe. Those advantages include features that are specifically
`applicable to diphtheria toxins such as CRM197. Specifically, the author
`7
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`states, and I quote “the reducing environment of the cytoplasm does not
`allow the formation of disulfide bonds which can easily be formed in the
`oxidizing environment of the periplasm.” Now the author has noted other
`advantages as well, including generation of a more stable protein due to the
`lower proteolytic activity in the periplasm and the ability to release the
`secreted protein from the cell via osmotic shock. So that’s, this paper and
`many others cited in the record reflect that more than 20 years alleged
`invention of the 345 patent, a person understood that the periplasm of E. coli
`was an ideal environment for expression of diphtheria toxins including
`CRM197. So, the issues to be decided here is whether a polynucleotide that
`encodes CRM197 or CRM197 like proteins has a signal sequence not
`derived from diphtheria and capable of directing transport of that CRM197
`and CRM197 like protein to the bacterial host periplasm was obvious to a
`person of skill in the art as of October 2009 and the petitioner respectfully
`submits that such a polynucleotide was obvious as of that timeframe.
`Turning to slide 30. Here we summarize the two grounds at issue in
`this matter. Obviousness over Davis in view of Inouye. And obviousness
`over Zhou in view of Ikemura.
`Turning to the first ground, slide 32. Davis is a US Patent that is prior
`under 102E. Davis was filed in 2007 and was published in 2009. But Davis
`is an example of the state of the prior art at the time of the alleged invention
`in 2009. Davis discloses and discusses compositions of modified diphtheria
`toxin and fusion protein containing those modified diphtheria toxins that
`reduced binding to vascular endothelial cells. The objective of the work
`reflected in Davis was to reduce incidents of something called Vascular
`Leak Syndrome. Davis notes at paragraph 7 that Vascular Leak Syndrome
`8
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`can occur following administration of IL-2, growth factors, and antibodies
`and chemotherapy and (inaudible) and organ failure. So, the objective of the
`work reflected in Davis as stated in paragraph 8 of Davis, was to design
`modified diphtheria toxins that cause reduced vascular leak syndrome
`compared to Wild Type Diphtheria toxin. The Davis notes at paragraph 15
`that the modified diphtheria toxins have reduced binding activity to human
`vascular endothelial cells as compared to unmodified diphtheria toxin. At
`paragraph 57 Davis discusses something called ONTAK and states that
`ONTAK is a protein fusion toxin containing a diphtheria toxin, toxophore
`andInterlukin-2. And ONTAK targets Interlukin-2 bearing cells for the
`treatment of lymphoma. Now as petitioner’s expert Dr. Georgiou explains
`in exhibit 1074 paragraph 64, Davis discusses the construction of
`approximately 100- 250 diphtheria toxin variants. And testing those variants
`for reduced vascular leak syndrome activity. Now, with all of that context
`we turn to slide 33 here we have reproduced Davis figure 2. In Davis figure
`2 presents diphtheria toxins specific antibodies detection and illustrates the
`binding of CRM197 which Davis refers to as DT-GLU52 to human vascular
`endothelial cells.
`If you turn to slide 34, here we have reproduced Davis figure 3. Now
`figure 3 shows binding for ONTAK and CRM197 to human vascular
`endothelial cells. As you can see the ONTAK binds the cells more than the
`CRM197. If you move to slide 35. On slide 35 we have reproduced figure 5
`from Davis. And figure 5 shows cell structural changes such as loss of cell
`membrane integrity after treatment with ONTAK or CRM197. And Davis
`notes that ONTAK “appears to cause more membrane damage than”
`CRM197. If you turn to slide 36 here, we have example 9 of Davis. In
`9
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`example 9 is Gene Synthesis, Expression and Purification, and in this
`example, Davis states for generation of full length diphtheria toxin and
`diphtheria truncate, meaning one that has only two of the three domains of
`diphtheria toxins, no receptor binding domains. And Davis goes on, to be
`used in cellular cytotoxicity and human vascular endothelial cell binding
`assays so these full lengths and truncates are going to be used in these assays
`which it refers to as stage one of the work. Vector systems are used which
`include secretory leader sequences for export of the diphtheria toxins in the
`periplasm of E. coli. Now, here Davis is explicitly disclosing production of
`full length and truncate diphtheria toxins through expression in the
`periplasm of E. coli with the secretory leader sequence. Now if you turn
`to…
`
`JUDGE SMITH: Counsel I have a question
`MS. DEVINE: Sure.
`JUDGE SMITH: So, there was discussion in the briefing about
`CRM197 being used as a control. But is it the petitioner’s position that at
`least as illustrated in figure 2 or perhaps in another figure that its not really
`being used as a control as much as a possible modified DT protein?
`MS. DEVINE: Thank you, your honor. It is the petitioner’s position
`that it is being used as a control. If we go back to slide 34 and figure 3, here
`you can see ONTAK and CRM197, two versions of fluorescent labeled
`CRM197. And a close reading of Davis and Dr. Georgiou explains within
`his declaration shows that what’s happening here, Davis is looking at the
`binding affinities of ONTAK and the binding affinities of CRM197 and so
`here we have a ceiling and a floor. And then with looking at experimental
`construct, of truncates other modified diphtheria toxins and figuring out
`10
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`where they fall on this scale and that gave an understanding of what sort of
`symptoms of vascular leak syndrome they were going to show and that’s
`based on the known vascular leak syndrome activity of ONTAK and
`CRM197 which is reflected in figure 5.
`JUDGE SMITH: So, maybe I’m not sure what’s being meant by
`control. I think of a control as something where you already know how it
`works and you’ve done it many times. Just run it side by side something
`you are actually investigating. But, so is, I guess is CRM197 also being
`used for in a purpose for other than a control? Or not?
`MS. DEVINE: In this work, its purpose is as a control. And the work
`in Davis, what Davis did was Davis took truncates of diphtheria toxin and
`the goal was to put different, make different fusion proteins that could bind
`different types of cells and use it that way. So, the truncates were the actual
`experimental construct. The CRM197 was the baseline for looking at those
`experimental constructs. There is discussion in the papers and the patent
`owner raises that the figures don’t explicitly show data comparing CRM197
`vs experimental construct in the same figure. Dr. Georgiou points out in his
`declaration, and I can get a citation later if it would be helpful. Dr. Georgiou
`points out that there were 100 to 250 experimental constructs. And it was
`simply not reasonable to expect that in every single one of those runs
`CRM197 would be run side by side, but it doesn’t mean that wasn’t function
`of the way that it was being run.
`JUDGE SMITH: So, in paragraph 324 Davis is there reference to full
`length. Is that referring to CRM197 because that’s a full length DT protein?
`MS. DEVINE: Yes. In example 9 the reference to full length DT
`protein it is petitioners’ position that it is referring to CRM197 which is a
`11
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`full length DT protein it’s not a truncate. And there is in fact no other full
`length protein, DT protein mentioned in Davis. I suppose one could think it
`means Wild Type, but that doesn’t seem reasonable, because Wild Type
`would be toxic.
`JUDGE SMITH: Thank you.
`MS. DEVINE: Thank you.
`MS. DEVINE: Okay moving on to slide 37. So as noted earlier,
`Davis discusses his work in stages and we have attempted to distill that
`down here in slide 37. Example 8 and 9 Davis provide details of how the
`work was carried out. Now stage 1 is the assay for diphtheria toxin activity
`related to vascular leak syndrome which is depicted in figures 2, 3 and 5.
`Stage 2 is gene synthesis, expression and purification of full length
`diphtheria toxins and diphtheria toxin truncates using a leader sequence for
`secretions into the periplasm of E. coli. And stage 3 involves the generation
`of testing of variants for assessment of vascular leak syndrome activity…
`pardon me… now importantly Davis instructs that the variants must be
`reliably produced such as the activities can be reliably compared. And this
`goes to the conversation we just had, your honor, is that if you are going to
`compare them, they should be produced in the same manner. And again,
`turning to slide 38. Again we have example 9 here and here in example 9
`Davis states, the method developed in stage 2, which is the gene synthesis
`expression and purification of whole diphtheria toxin in E. coli, that’s what
`it includes; provides for reliable production of multiple diphtheria toxins
`variants with similar quality such that the activities of these variants can be
`accurately compared. If we turn to slide 39. Actually, let’s just skip to slide
`40. Here we have reproduced example 8 and figure 3 from Davis. Now
`12
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`Davis states that the CRM197 was used as a control here. If you look in
`paragraph 308, which is on this slide, it states the assay achieved good
`detection of ONTAK and controlCRM197, and that’s fluorescent labeled
`CRM197. Now we turn to slide 42.
`JUDGE SMITH: Counsel, counsel before you go ahead, from slide
`40. So, I had some issues with your reliance on figure 3 because its not
`simply referring to DT protein. Could you respond to that please?
`MS. DEVINE: Yes. Sure, so you’ll see its labeled DT-GLU52 which
`is CRM197- and then there is A1488. What that references the fluorescein
`label which is how it was measured. The CRM197 was labeled and that’s
`how they look at it. So, moving to slide 42 and Inouye. So, Davis explicitly
`references the paper Inouye 1985, and says that the pin vectors disclosed in
`that paper are useful in the work that Davis discusses. Now that 1985
`Inouye paper discusses among other things pIN-III-ompA and stated that a
`facilitated secretion of protein in the periplasmic space of E. coli.
`Now turning to slide 43. As Dr. Georgiou explains in exhibit 1074
`paragraph 70-71, the work of Inouye conducted in the 1970’s and 80’s was
`groundbreaking, and a person of skill in the art would have been well aware
`of it by 2009. Turning to slide 44, also as Dr. Georgiou explains this Inouye
`paper, this 1985 paper explicitly sited by Davis is only one of many
`references published by the Inouye laboratories that discusses the expression
`of heterologous protein in E. coli. Dr. Georgiou further notes that Dr.
`Inouye generously distributed his pin-III-ompA vectors resulting in labs
`including Dr. Georgiou’s producing heterologous proteins in the periplasm
`of E. coli. Many such proteins are summarized in our exhibits, exhibit 1030
`at table 5. So as summarized earlier, the challenged claims cover a
`13
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`polynucleotide that has two requirements. 1. A portion that encodes a
`mature polypeptide having a sequence that is at least 90 percent identical to
`CRM197. 2. A portion not derived from diphtheria that encodes a
`polypeptide capable of directing transport of that toxin peptide to the
`periplasm of the host cell. Last discussed, Davis discloses expression of full
`length diphtheria toxins including CRM197 and diphtheria toxins truncates
`in the periplasm of E-coli and states that the pin vectors of the 1985 Inouye
`paper which includes sequences that are not derived from diphtheria
`including pIN-III-ompA are useful in the work discussed in Davis. Now
`accordingly the challenged claims would have been obvious to a person of
`skill in the art by October 2009. Moreover, the proposed amended claims do
`not alter this conclusion. They do not change the structure of the claimed
`polynucleotide. The patent owner has not presented any argument to the
`contrary. If we move to ground 2, which starts on slide 46. Ground 2 is
`Zhou in view of Ikemura. So, the Zhou paper published in 1999, so this is
`earlier than the Davis paper. And Zhou discussed expression of CRM197
`under the subtilisin sequence, the signal sequence for subtilisin in gram
`positive bacteria. Now Zhou notes at page 255 that the yield of the secreted
`protein in that Gram-positive bacteria was less than what was seen in E. coli.
`And in the abstract, Zhou also notes that the secreted protein underwent
`degradation. If you turn to slide 48, here we have reproduced figure 1 of
`Zhou. And as Dr. Georgiou explains, that exhibit 1002, paragraph 183 and
`exhibit 1074 paragraph 53 and 54, Gram-positive bacteria such as the B.
`subtilis used in Zhou, they do have a periplasmic space but its small. So
`typically, what is done is proteins that are expressed in Gram-positive
`bacteria are expressed directly under the culture medium. Now because they
`14
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`are expressed in the culture medium, they frequently undergo degradation.
`An as it can be seen here in Zhou figure 1, the protein that was expressed in
`this paper did undergo such degradation. Now if we turn to slide 49, as Dr.
`Georgiou further explains because of that degradation, Gram-positive
`bacteria were not widely accepted as host bacteria by persons of skill in the
`art in 2009. This reality is reflected at least in this paper which is exhibit
`1041 which is excerpted here this is a 2005 paper that is co-authored by Dr.
`Georgiou. And in that paper, the author states that Gram-Positive bacteria
`have not achieved widespread acceptance as hosts for the expression of
`secreted proteins, largely because the organisms so far examined, including
`B. subtilis, excrete high levels of proteolytic enzymes that can cause
`extensive product degradation. If you turn to slide 50, here we have
`excerpted Ikemura, which is the second reference in ground 2. Ikemura was
`published in 1987. And Ikemura discusses expression of heterologous
`protein in the periplasm of E. coli, Gram-negative bacteria and based on the
`understanding of the disadvantages as explained by Dr. Georgiou of gram-
`positive bacteria, a person of skill in the art in 2009 would understand that E.
`coli as discussed in Ikemura was a superior host cell. If you turn to slide 51,
`now Ikemura discusses expression of (inaudible). Ikemura used both the
`native signal sequence and an E. coli derived OmpA signal sequence. Thus,
`Ikemura demonstrated in 1987, more than 20 years before the supposed
`invention of the 345 patent that a protein could be secreted in the periplasm
`of E. coli using either the subtilis signal sequence which is what Zhou used
`to track CRM197, or the ompA signal sequence. So as summarized earlier,
`2 pieces of the claim, the portion that encodes the mature polypeptide having
`the sequence that is at least 90 percent CRM197 and the portion not derived
`15
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`from diphtheria that is the signal sequence capable of directing the toxins.
`Zhou discloses such a polynucleotide. Zhou discloses a polynucleotide
`encoding a subtilis signal sequence and CRM197, that is the polynucleotide.
`Ikemura demonstrates that such a signal sequence in fact is capable of
`directing transport into periplasm that in view of Zhou and Ikemura are of
`skill in the art as of 2009 would find the claim polynucleotide obvious and
`again it does not change the structure of the polynucleotide the amended
`claims do not change that conclusion.
`JUDGE SMITH: Counsel, patent owners have taken the position that
`Zhou actually teaches a way from periplasmic expression secretion. Can
`you speak to that please?
`MS. DEVIN: Sure. So, Zhou…let me just open my copy of
`Zhou…Zhou does not teach a way. Zhou does not say not to use E-coli.
`Zhou presents B. subtilis as in Zhou’s view an acceptable expression system.
`But Zhou does not say, do not use E. coli which is what would be required
`for an actually teaching a way. What Zhou does state is that the amount of
`protein obtained and this is at Zhou 255; Zhou states that the yield of the
`protein obtained were not so high as that in E. coli. Zhou had hypothesized
`that there are things that can be adjusted to improve yield. So, Zhou actually
`points out that yield any E. coli is better than B. subtilis, it’s certainly not
`teaching away.
`Now if we move back to slide 6. Here we’ve outlined the claim
`construction dispute. And petitioner respectfully submits that patent owner
`claim construction dispute is an attempt to obfuscate the clear
`unpatentability of the claim. If you turn to slide 9 here, we have excerpted
`some of the patent owners’ arguments from the papers. And patent owner’s
`16
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`interpretation of this dispute frame things as whether the claim nucleotide
`actually does or does not perform an action. And that simply makes no
`sense because these are not method claims they are not actionable steps.
`And the Federal Circuit addressed this issue in ParkerVision v. Qualcomm in
`2018. And the Federal Circuit stated, apparatus claims covered what a
`device is not what a device does. Here the 345 claims cover what the
`polynucleotide is not what the polynucleotide does. The Federal Circuit
`went on in ParkerVision and said an apparatus that is quote capable of
`performing a certain function may be rendered obvious in view of a prior art
`apparatus that can likewise perform those functions. That is the case here.
`Here the claimed polynucleotide is obvious in view of prior art
`polynucleotides that could direct transport of the bacterial toxin polypeptides
`to the bacterial periplasm when expressed in a host cell.
`JUDGE SMITH: Counsel
`MS. DEVINE: Yes, your honor?
`JUDGE SMITH: Help me out here. So, I’m sort of with you in the
`sense that, obviously there is a claim to a polynucleotide but its described in
`functional terms. Both the 3 prime sequence and the 5 prime sequence do
`certain encoding and that’s their function. Their function is to encode.
`Now, is it the petitioner’s position that those functional limitations are not
`limitations? Or are they limitations, that somehow define the structure or
`describe the structure?
`MS. DEVINE: Hopefully I understand your question your honor.
`And let me know if I am not being responsive. We are not arguing that that
`language is to pertinent to within the claims, that is not our argument. The
`actual polynucleotide has to be capable of as in ParkerVision stated the
`17
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`apparatus has to be capable of, its not meaningless language. But the actual
`thing that is claimed, is that the actual polynucleotide. And what we have to
`show as the petitioner, is that prior art rendered obvious that polynucleotide.
`We don’t have to show, which is what I suspect the patent owner is arguing
`from the papers, we don’t have to show that a person skilled in the art
`viewing, viewing the prior art would have some certainty that a given
`polynucleotide would absolutely succeed under any circumstance in
`expressing the protein in the periplasm. It has to be capable of doing that
`but there doesn’t have to be any sort of certainty. Is that responsive to your
`question?
`JUDGE SMITH: Yes, thank you.
`MS. DEVINE: Alright. Moving on to slide 19. So, the second issue
`that’s been raised in the record is the state of the prior art. The patent owner
`attempts to confuse the issue by essentially citing outdated prior art that
`shows alleged challenges and problems in periplasmic expression of
`diphtheria toxins. Now on slide 19, we summarize some of these old
`publications cited by patent owner’s expert Dr. Galen, which dates 10 and
`25 years before the alleged invention. Now if you turn to slide 21, we
`excerpted Dr. Galen’s deposition testimony where Dr. Galen admitted that
`those references are not contemporary with the 2009 potential priority dates
`of 345 patent. Now if you turn to slide 26, on slide 26 and 27, we
`summarize several examples of disclosures of polynucleotides capable of
`expression of full length diphtheria toxin and diphtheria toxin variance in the
`prior art. So, beginning with the first, Bishai, which is exhibit 2010, the
`1987 paper and it discloses periplasmic expression in e-coli of CRM197 and
`notes that there was minimal proteolysis. We’ve already discussed Zhou
`18
`
`
`
`1
`2
`3
`4
`5
`6
`7
`8
`9
`10
`11
`12
`13
`14
`15
`16
`17
`18
`19
`20
`21
`22
`23
`24
`25
`26
`
`

`

`IPR2019-01028
`Patent 9,422,345 B2
`
`
`and Collier which is exhibit 1050 which issued in 1999 it states, and this is
`at column 6 lines 33-40, DNA encoding diphtheria toxin or a polypeptide of
`the invention may also be carried on any other vector operably linked to
`control signals capable of effecting expression in the prokaryotic host. If
`desired the coding sequence can contain at its five prime end, a sequence
`encoding any of the known signal sequences capable of effecting secretion
`of the expressed protein into the periplasmic space of the host cell, thereby
`facilitating recovery of the protein. Another example Mekada, said that
`1049 said that in 2006 discussed cloning CRM197 into a plasmid vector that
`included a non-diphtheria signal sequence for periplasmic secretion in E.
`coli. As we have discussed Davis, also discusses periplasmic secretion of
`full length diphtheria toxins in E. coli. If you turn to slide 27, Zettlemeissl
`exhibit 2010, in 1986 discussed production of full length diphtheria mutant,
`this one’s known as CRM228. And then we also cite 2 other papers exhibits
`1069 and 1070 that also discuss expressions of those same diphtheria
`mutants. So, petitioner respectfully submits the patent owner’s argument
`that there is somehow a deft of prior art reflecting successful expression in
`the periplasm of gram negative bacteria of diphtheria variants or mutants is
`simply not true.
`And moving on to patent owners’ motion to amend, I’ve already
`touched on it some but just to sum it up, on slide 53 we’