[00:00:00] Speaker 02:
Our final case this morning is number 22, 1946, Pure Circle USA, Inc.

[00:00:06] Speaker 02:
versus Shujen, Inc., Mr. Panachowski.

[00:00:14] Speaker 03:
Yes, Your Honor.

[00:00:15] Speaker 03:
Thank you, Your Honors.

[00:00:16] Speaker 03:
Stanley Panachowski for the appellants, Pure Circle.

[00:00:20] Speaker 03:
The district court's judgment should be reversed because it erred as a matter of law on both written description and on section 101.

[00:00:30] Speaker 03:
On written description, the court improperly resolved disputed issues of material fact about the sufficiency of the patent's disclosure.

[00:00:39] Speaker 02:
Okay, help me understand what's going on here.

[00:00:42] Speaker 02:
As I understand it, there are either four or five enzymes that have been identified.

[00:00:51] Speaker 02:
that would achieve this conversion, correct?

[00:00:56] Speaker 02:
Correct, Your Honor.

[00:00:57] Speaker 02:
And you have said in several places in your brief that there are only about 1,800 possibilities, but

[00:01:07] Speaker 02:
I'm wondering whether that's really correct, because as I read the record, the 1800 possibilities relate to a single enzyme, not to the four or five, correct?

[00:01:17] Speaker 02:
Correct, Your Honor.

[00:01:19] Speaker 02:
And then there is the possibility that in the future, that additional enzymes would be identified, which achieve the conversion, and the claims would cover those new enzymes as well, right?

[00:01:32] Speaker 02:
Correct, Your Honor.

[00:01:33] Speaker 02:
So why isn't that

[00:01:35] Speaker 02:
problem from a written description standpoint.

[00:01:37] Speaker 02:
If you'd wanted to

[00:01:39] Speaker 02:
to limit your universe, the scope of the claim, to the four or five existing enzymes, you could have easily said that and look at claim 14, which is limited to one of the enzymes, but you chose to write it broadly so that it covers enzymes that have not been discovered yet and may be discovered in the future.

[00:02:00] Speaker 02:
So that's a fairly broad claim and that in itself, it seems to me, creates a written description problem.

[00:02:09] Speaker 03:
Your Honor, there is no written description problem here.

[00:02:12] Speaker 03:
And I will walk through those numbers to show why, when you view this specification through the lens of a skilled artisan's knowledge and the prior art, as this court instructed in Ariad, it's sufficient disclosure for a reasonable jury to say, Sweden hasn't met their burden by clear and convincing evidence.

[00:02:31] Speaker 03:
So first, Your Honor, as you said,

[00:02:33] Speaker 03:
Claim 14 is different here because it specifically identifies the UGT76G1 enzyme.

[00:02:41] Speaker 03:
And as your honor also said, the 1800 number does relate to that particular enzyme.

[00:02:49] Speaker 03:
And what is most important for the other claims about that 1800 number is to look at the methodology.

[00:02:57] Speaker 02:
But you drafted those claims to cover

[00:03:00] Speaker 02:
not only all the known enzymes, but any future enzymes as well, right?

[00:03:06] Speaker 03:
Yes, your honor, but there is no evidence in the record that skilled artisans have any reason to expect that there are going to be future unknown enzymes that lack the common structural features or the structure-function correlation that we see in the record here.

[00:03:24] Speaker 02:
What we know from... Well, what are the structural features that we're talking about?

[00:03:29] Speaker 02:
Common structural features?

[00:03:31] Speaker 03:
Yes, Your Honor.

[00:03:32] Speaker 03:
The common structural features are the specific folds and helices that are shown in the model of the UGT76G1 enzyme, for example, in the Olson article and also cited by Pura Circle's expert.

[00:03:49] Speaker 02:
If I understand correctly, those

[00:03:52] Speaker 02:
structural features are peculiar to that one enzyme.

[00:03:55] Speaker 02:
They're not structural features that are necessarily shared with other enzymes that achieve the conversion.

[00:04:03] Speaker 03:
No, Your Honor, because there is evidence in the record both from the 2005 Richmond article and the 2009 Keyshore patent application

[00:04:14] Speaker 03:
Both are cited in the patent and discussed extensively by Dr. Bollinger.

[00:04:19] Speaker 03:
And what those references, known by a skilled artisan, show is that even though that very particular three-dimensional structure relates to UGT76G1, it was known in the art that all of the claimed UDP-glucosal transferases

[00:04:38] Speaker 03:
had certain common structural features.

[00:04:40] Speaker 03:
And those are identified in that figure three from the Olson article that Dr. Bollinger cites.

[00:04:47] Speaker 02:
And what we see in there.

[00:04:48] Speaker 02:
There's a testimony that all the known enzymes had common structural features.

[00:04:54] Speaker 03:
Your Honor, that testimony is in Dr. Bollinger's rebuttal expert report at appendix pages 5568 and 5583.

[00:05:02] Speaker 03:
5568?

[00:05:05] Speaker 02:
Yes, Your Honor.

[00:05:11] Speaker 02:
And 5583 was the other one?

[00:05:14] Speaker 03:
Yes, that was the endpoint, Your Honor.

[00:05:16] Speaker 02:
So it's 5568 through 5583.

[00:05:20] Speaker 02:
Yes, Your Honor.

[00:05:21] Speaker 02:
But where does he say this?

[00:05:23] Speaker 02:
5568.

[00:05:25] Speaker 03:
What we see here is that at 5569, he's talking about the Richmond article.

[00:05:32] Speaker 02:
Wait a moment.

[00:05:33] Speaker 03:
Yes, Your Honor.

[00:05:34] Speaker 02:
Is that in the appendix?

[00:05:36] Speaker 03:
Yes, Your Honor, it is.

[00:05:38] Speaker 03:
There's also another stretch where the same pages appear, but this is in the appendix.

[00:05:45] Speaker 02:
My appendix goes from 5559 to 6323.

[00:05:47] Speaker 02:
5569?

[00:05:59] Speaker 02:
Those pages seem to be missing from my appendix.

[00:06:05] Speaker 03:
Your Honor, if you have

[00:06:06] Speaker 03:
Appendix page 6430.

[00:06:08] Speaker 03:
The same information is here because the parties submitted the same exhibit multiple times in connection with the summary judgment motion.

[00:06:19] Speaker 03:
So you can also find the exact same material started at Appendix page 6430.

[00:06:25] Speaker 01:
What's the best point?

[00:06:28] Speaker 01:
Maybe a sentence somewhere where Dr. Bellinger says... I don't have those pages in my appendix either.

[00:06:34] Speaker 03:
Those are in volume two, Your Honor.

[00:06:36] Speaker 02:
I'm in volume two.

[00:06:38] Speaker 02:
They're not there.

[00:06:39] Speaker 02:
I mean, it really is incumbent upon you to make sure that the documents that you submit to us are complete.

[00:06:52] Speaker 02:
OK, go ahead.

[00:06:53] Speaker 01:
On the structure, that the structure was known, where can we find that most efficiently?

[00:07:00] Speaker 03:
Yes, Your Honor.

[00:07:01] Speaker 03:
The pivotal sentence would be on 5570 or 6431, where Dr. Bollinger says, after talking about the Richmond article, paragraph 286.

[00:07:15] Speaker 03:
So after talking about the Richmond article, which identified the 12 enzymes and then talked about the five functional ones, because they all have this signature amino acid sequence, Dr. Ballinger says, there may be others, but the number known in 2012 was small, and they all had common structural features.

[00:07:36] Speaker 02:
Where is this at?

[00:07:37] Speaker 02:
5570?

[00:07:37] Speaker 03:
5570 and 6431.

[00:07:40] Speaker 02:
But he's not timing the common structural features.

[00:07:43] Speaker 02:
to the active regions that do the conversion.

[00:07:47] Speaker 02:
I mean, if that's all he says, that's pretty vague.

[00:07:51] Speaker 02:
Your Honor, that is not all he says.

[00:07:53] Speaker 02:
That's the sentence.

[00:07:55] Speaker 02:
Where does he say that it has common structural features in the active regions?

[00:07:59] Speaker 03:
Correct.

[00:08:00] Speaker 03:
So following that, we look at page 5571, which is also at page 6432.

[00:08:08] Speaker 03:
And that's the rest of the range here.

[00:08:10] Speaker 03:
your honor, where he is showing this model and pointing out the particular active sites where you have your donor molecule docking.

[00:08:22] Speaker 02:
That's the... You're interested that the active sites would be structurally identical.

[00:08:26] Speaker 03:
On page 5572 you have figure three there and then he explains right below that, that here we can see the red M, which is also red X in the claims, sitting in the pocket and can tell what amino acids in the active site make crucial interactions with the substrate.

[00:08:46] Speaker 02:
But he's not saying that other enzymes that perform the conversion have the same structure.

[00:08:52] Speaker 03:
Your Honor, he's not saying that, but here

[00:08:55] Speaker 02:
sweeten has the burden of proof by clear convincing evidence and we're on summary judge of the i'd thought you'd suggested that there was uh... testimony that there was common structural features well version and now it seems you're admitting there is no testimony that there is common structural features no your honor know what no no i'm not admitting that there's a testimony the testimony that we're looking at is the case only that says that someone that but doesn't say that

[00:09:26] Speaker 03:
Your Honor, it does say that because what Dr. Bollinger is doing, and again, it's not just Dr. Bollinger, it's pointing to the Richmond article, Keyshore, Swegins, experts Dr. Olson's work, and what he's saying is these common structural features, we see where the UDP glucose donor molecule is sitting.

[00:09:49] Speaker 02:
We also see where the... Where does he say that those are common structural features?

[00:09:55] Speaker 03:
Your honor, he says that these are common structural features in paragraph 286, and that's the beginning of his description of what these common structural features are.

[00:10:09] Speaker 02:
I'm sorry, your honor?

[00:10:12] Speaker 02:
Paragraph 286 is the only place that he talks about common structural features.

[00:10:16] Speaker 03:
No, Your Honor, that is essentially the topic paragraph for everything that follows here.

[00:10:22] Speaker 03:
And what follows is the description of the common structural features that are mentioned in an introductory way.

[00:10:29] Speaker 02:
Where in your brief did you argue that these enzymes that perform this conversion all have common structural features?

[00:10:37] Speaker 03:
Your Honor, we argued this throughout both our opening briefs.

[00:10:42] Speaker 03:
Just give me a page where you argue that.

[00:10:48] Speaker 03:
So Your Honor, starting with the reply brief.

[00:10:53] Speaker 03:
We're in the opening.

[00:10:55] Speaker 03:
Yes.

[00:10:55] Speaker 03:
I'll go back to the opening brief.

[00:11:03] Speaker 03:
So Your Honor, starting at page 24, this is the session that contains that argument.

[00:11:09] Speaker 03:
And then on page 26, we show that Dr. Bollinger cited.

[00:11:15] Speaker 03:
Just show me.

[00:11:17] Speaker 03:
Show me the sentences.

[00:11:18] Speaker 03:
Yes, Your Honor.

[00:11:19] Speaker 03:
Page 26 of the opening brief.

[00:11:21] Speaker 03:
More specifically, Dr. Bollinger cited prior evidence showing that in 2012, a poster would understand that UDP glucosal transferases have a common folded structure and have common amino acid active sites that must be conserved, i.e.

[00:11:38] Speaker 03:
not eliminated across mutant variants, to maintain the glucose transfer function of the claims.

[00:11:45] Speaker 03:
and it cites pages from the excerpt we were just discussing.

[00:11:48] Speaker 03:
And then the brief says that Dr. Bollinger identified prior art showing that these components must be physically held in a particular proper orientation to be able to complete the addition of the activated glucose unit.

[00:12:02] Speaker 03:
And then he goes on to

[00:12:05] Speaker 03:
discuss the evidence that the UDP Glucosal Translators have an identifying signature amino acid sequence.

[00:12:11] Speaker 03:
And then there is the use of that three-dimensional model, which is also discussed further at pages 50 of the opening brief and forward, where you then see how this information known in the prior art was used to construct these three-dimensional models.

[00:12:32] Speaker 03:
The main issue here, especially with respect to Claim 14, is mutants.

[00:12:37] Speaker 03:
So what do we do with mutants?

[00:12:38] Speaker 03:
Well, this court's decision in Ajinomoto provides a blueprint for how you handle mutants.

[00:12:44] Speaker 02:
The way I read this, and maybe I'm reading it incorrectly, is that mutations of a particular enzyme would have common structural features, not that different enzymes have common structural features.

[00:12:58] Speaker 03:
Both are true, Your Honor.

[00:13:02] Speaker 01:
Can I ask you, it seemed like the district court's main concern was with the breadth of the claim construction that you agreed to, and that that's what opened up possibly trillions of embodiments that I think even you would have to agree there's not adequate written description for trillions of embodiments.

[00:13:19] Speaker 01:
So help me understand, is this claim construction, is it a definition from the specification, or is it just you made a mistake and agreed to a claim construction that for whatever reason there's not adequate written description?

[00:13:32] Speaker 03:
Your Honor, so the claim construction, which affects every claim except for claim 14 of the 273 patent, that itself is not a purely functional definition because it comes from the specification in column... So it does come from the specification.

[00:13:50] Speaker 03:
Yes, but Your Honor, with one change, where the specification said that a UDP-glucosal-transferase enzyme is

[00:14:00] Speaker 03:
Type of enzyme that then performs or is any can be any UDP?

[00:14:05] Speaker 03:
Glucosal transferase that performs a transfer function the stipulated claim construction Doesn't repeat the word UDP?

[00:14:13] Speaker 03:
Glucosal transferase in that definition rather it says a type of enzyme that performs, and that's very broad Yes, your honor, but it is not as broad as any enzyme

[00:14:25] Speaker 03:
which is how it has now been interpreted by the district court in its original claims doctrine analysis as well as switching.

[00:14:33] Speaker 02:
But it's not any enzyme.

[00:14:35] Speaker 02:
It's any enzyme that performs the conversion, right?

[00:14:39] Speaker 03:
Your Honor, the stability claim instruction uses the language a type of enzyme.

[00:14:44] Speaker 03:
And when you look at that in light of the specification passage from which it borrows, it's not trying to eliminate

[00:14:50] Speaker 03:
the notion of UDP-glucosal transferase being a structural limitation on the class.

[00:14:56] Speaker 03:
However, Your Honor, even if one accepted that it could include any enzyme and that it's a purely functional definition, the district court's own opinion shows that's not the end of the analysis, it's the end on original claims doctrine, but it's the beginning of the analysis

[00:15:12] Speaker 03:
on the three tests that this court applied in Juneau.

[00:15:16] Speaker 03:
And what the evidence here shows is that just taking one of those three tests, the structured function correlation test, that there were tools known to one of skill in the art, disclosed in prior art, discussed in the patents, acknowledged by Sweeten's own experts, from which you are not left like a skilled artisan in Juneau.

[00:15:35] Speaker 03:
to flounder, not knowing how to distinguish which mutants would work and which wouldn't, you could use homology modeling to narrow it down and then routine automated assay testing to get your definite number of mutants.

[00:15:48] Speaker 02:
But as I understand it, that homology modeling would not help you identify other enzymes, not mutations, but other enzymes beside the four or five that perform the conversion, correct?

[00:16:01] Speaker 03:
Correct, Your Honor.

[00:16:02] Speaker 03:
If there were other enzymes that lacked those structural features, here Swigin has the burden of proof.

[00:16:09] Speaker 03:
And Swigin did not identify in this record a single example of an enzyme that could perform the function that lacked those common structural features.

[00:16:20] Speaker 03:
And absent that evidence,

[00:16:24] Speaker 02:
Hypothetically, I know you disagree with this, but hypothetically that we think there is no evidence on your side of common structural features among different enzymes that perform the conversion.

[00:16:39] Speaker 02:
Is then the district court correct that there's no written description support?

[00:16:44] Speaker 03:
no your honor even if we had an absence of evidence we know from the supreme court's decision in seal attacks that's region is the party bearing the burden of proof at trial and hereby clear convincing evidence but they have the burden prove but don't you have a burden of production to identify common structural features if they exist your honor we have that burden of production only if

[00:17:10] Speaker 03:
They meet their initial burden to show that they have evidence from which a reasonable jury could conclude.

[00:17:19] Speaker 03:
Yes, by clear and convincing evidence, there are no common structural features, no structure function correlation, no representative species.

[00:17:27] Speaker 03:
And Your Honor, this is operating under your hypothetical.

[00:17:30] Speaker 03:
As we've discussed today, there's ample evidence that pure circle produce

[00:17:36] Speaker 03:
that shows that you can start with what's identified in the specification with these enzymes and these specific sequences for the genes that you use to encode them to express these enzymes.

[00:17:48] Speaker 03:
Then you go to the prior ARC cited in the patent to deal with mutants.

[00:17:54] Speaker 03:
And we can deal with mutants with homology modeling and then automated assay testing.

[00:17:59] Speaker 03:
And then you're not going to be in the Juno bucket, where you simply have no way to distinguish between what's claimed and what's not.

[00:18:07] Speaker 03:
Instead, you're going to be in the Ajinomoto bucket, where this court acknowledged, yes, there is evidence of counterexamples here, that there's a record that says when you have mutations toward the consensus sequence, generally that increases promoter strength and vice versa.

[00:18:23] Speaker 03:
And sure, the defendant pointed to articles that said, that's not always the case.

[00:18:27] Speaker 03:
deviations here toward the consensus sequence that have weaker function, and then we have deviations away from them that have stronger function, and this is where it's at.

[00:18:35] Speaker 01:
Is there anywhere in either the briefing or the appendix where you list out what you think are the genuine disputes of material fact that made it premature to grant summary judgment?

[00:18:48] Speaker 03:
Your Honor, there are portions in the appendix that do cite to the statement of undisputed and disputed material facts that were identified below.

[00:18:57] Speaker 03:
I think that the best source for that, Your Honor, would be in the reply brief, which is referring back to arguments made in the opening brief.

[00:19:06] Speaker 03:
What pages?

[00:19:07] Speaker 03:
Yes.

[00:19:07] Speaker 03:
So Your Honor, if you look at the table of contents,

[00:19:12] Speaker 03:
Under the argument section, in section one, we laid out what we think the most important, genuinely disputed facts are here.

[00:19:22] Speaker 03:
And it's the structure of UDP, glucozol transfers was known at the time of filing, correlation between structure and function,

[00:19:29] Speaker 03:
talking about fusions, mutants, et cetera.

[00:19:32] Speaker 03:
And then in section two, we take each of these disputes of fact and show why it is dispositive here under this court's precedent that- So one A and B in your reply brief would be a good summary of where you think the genuine disputes of material fact are?

[00:19:48] Speaker 03:
Yes, Your Honor.

[00:19:49] Speaker 03:
And those disputes of material fact

[00:19:51] Speaker 03:
at least align this case with a genomoto, where the court acknowledged, just like Suede's evidence here, there is evidence going in the other direction, that you're not going to have a one-to-one perfect correspondence between structure and function.

[00:20:07] Speaker 03:
But this court said,

[00:20:09] Speaker 03:
We have never demanded that.

[00:20:11] Speaker 03:
Instead, we more modestly spoke an aria of a correlation, which means you can have counter examples.

[00:20:16] Speaker 03:
You might not have perfect predictions from structure.

[00:20:20] Speaker 03:
But if you have the tools, like the tools in the prior art and that Dr. Bollinger tested here, that enable you to narrow the universe of relevant

[00:20:31] Speaker 03:
and we're a skilled artist and can readily determine what works, you've done enough to get to the jury.

[00:20:35] Speaker 02:
Wait a moment.

[00:20:36] Speaker 02:
I'm confused.

[00:20:37] Speaker 02:
Now you're talking about using homology modeling to identify common features, which is different from identifying common structure.

[00:20:48] Speaker 02:
It's a different argument, right?

[00:20:50] Speaker 03:
No, Your Honor.

[00:20:51] Speaker 03:
Homology modeling has two roles in this technology.

[00:20:55] Speaker 03:
So the first thing that homology modeling does is we take an enzyme just like

[00:21:00] Speaker 03:
the Keyshore prior art did, or like Dr. Olson did.

[00:21:04] Speaker 02:
So are you suggesting that the structure was not known but could have been determined by homology modeling?

[00:21:14] Speaker 03:
No, Your Honor.

[00:21:14] Speaker 03:
In 2012, even though the crystal structure

[00:21:19] Speaker 03:
of the UDP glucosal translators wasn't known.

[00:21:22] Speaker 03:
The secondary structure, which is that three-dimensional model that we were talking about, you are able to determine that.

[00:21:29] Speaker 02:
So you're saying you don't need homology modeling because the common structure was known at the time.

[00:21:35] Speaker 03:
The homology modeling had already been done to produce the common structural features of the enzyme.

[00:21:42] Speaker 03:
And then the second way in which you use the homology model in your honor is then you take that model, for example, of the UGT76G1 enzyme and then you run it through the widely available computer software that then does modeling of potential mutants where you change amino acids at various points and then

[00:22:05] Speaker 03:
the computer program will tell you these are the mutations that can serve the structure that's needed to hold the right molecules in the right orientations.

[00:22:16] Speaker 02:
And that's how you get $1,800.

[00:22:16] Speaker 02:
We're out of time, but I'd like you to address briefly the 101 question, particularly in relation to claim 14, which does identify a specific enzyme.

[00:22:32] Speaker 02:
So why is the district court wrong about claim 14 being 101 and eligible?

[00:22:41] Speaker 03:
Your Honor, the district court is wrong about claim 14 because of the conversion percentage limitation that is in claim 1.

[00:22:50] Speaker 03:
And that conversion percentage limitation

[00:22:53] Speaker 03:
we know from the record, including admissions of Sweden's own experts, is not something that occurs in nature.

[00:23:00] Speaker 03:
Rather, as the examiner recognized in requiring Pure Circle to add that limitation to overcome a Section 101 rejection, in nature you don't get

[00:23:10] Speaker 03:
conversion percentages of at least about 50% or more.

[00:23:15] Speaker 03:
It happens only in the last.

[00:23:16] Speaker 02:
But I guess the argument would be that 50% is just a goal.

[00:23:20] Speaker 02:
There's nothing in the claims specification that tells you how to achieve the 50%, correct?

[00:23:28] Speaker 03:
Your Honor, there is no description in the claim of a process that distinguishes between getting the 50% and not what the process is to contact.

[00:23:40] Speaker 03:
these substrates.

[00:23:41] Speaker 03:
And then there is ample guidance in the specification as to actually how you put that together.

[00:23:46] Speaker 03:
The reason, Your Honor, why this doesn't fall under this Court's precedence, like American Axel talking about only a result, is here we have method of preparation claims, like in Illumina and Cells Direct.

[00:23:58] Speaker 03:
And what the Court has said is we look at these claims and we see, are they directed to a natural phenomenon?

[00:24:06] Speaker 03:
And here,

[00:24:07] Speaker 03:
Whatever arguments that Suijin wants to make, and that they've already made about Section 103, Section 112, about whether these result limitations help make the claim valid, in terms of eligibility, this court's precedents are dispositive because we know from the record you cannot achieve those conversion percentages in nature.

[00:24:31] Speaker 01:
What about Ariosa?

[00:24:33] Speaker 01:
Didn't we say something similar would not be patent eligible?

[00:24:38] Speaker 03:
No, Your Honor.

[00:24:39] Speaker 03:
Arioso wouldn't be similar, because that was not a method of preparation claim.

[00:24:43] Speaker 03:
When you have diagnostics claims, or perhaps claims that are drawn to something like a primer itself, this court has said

[00:24:53] Speaker 03:
that when the claim is directed to a natural law, the mere fact that you might be using some human-made tool to conduct that diagnosis doesn't automatically convert it into eligible subject matter.

[00:25:09] Speaker 03:
It'd be one thing if you're creating a new measurement tool and you're claiming that.

[00:25:14] Speaker 03:
That could be eligible.

[00:25:14] Speaker 03:
But if all you're doing is taking conventional techniques, whether they're natural or human-made,

[00:25:20] Speaker 03:
and then applying them to observe the natural law, those claims are generally not eligible.

[00:25:26] Speaker 03:
But this court has said when we're in the method of treatment and method of preparation universe, we look at it differently.

[00:25:32] Speaker 03:
And the beer.

[00:25:35] Speaker 02:
All this says, Plan 14, is a method that achieves a 50% conversion rate, right?

[00:25:41] Speaker 02:
Correct, Your Honor.

[00:25:42] Speaker 02:
And it doesn't tell you how to do it.

[00:25:44] Speaker 03:
Other than what is described earlier in that method, which is that you take these substrates

[00:25:50] Speaker 03:
and then you contact them with each other.

[00:25:52] Speaker 03:
That part of it was known in nature, right?

[00:25:56] Speaker 03:
Yes, Your Honor, except once you introduce the conversion limitation, you know that you're in the lab and this entire claim has to be performed by a human being because, again, we know from the record

[00:26:10] Speaker 03:
pages 6609 from Dr. Bollinger's testimony to experts who testify for sewage.

[00:26:16] Speaker 03:
And we know that these rates don't occur in nature.

[00:26:21] Speaker 03:
And therefore, we're in a situation also like natural alternatives, where in natural alternatives the court said, yes,

[00:26:28] Speaker 03:
These are all natural products that are being used in the claim.

[00:26:33] Speaker 03:
But we're treating a patient with unnatural amounts of the beta-alanine.

[00:26:38] Speaker 03:
And those amounts of this product do not occur in nature, even though the product itself doesn't.

[00:26:45] Speaker 03:
This isn't a method treatment, right?

[00:26:48] Speaker 03:
Correct, Your Honor.

[00:26:48] Speaker 03:
And that's why Illumina installs direct

[00:26:51] Speaker 03:
are the most directly factually on point precedents because they are talking about methods of preparation, distinguishing them from

[00:27:00] Speaker 03:
from the diagnostic claims.

[00:27:01] Speaker 03:
And what the court said in Illumina is here we have human engineered parameters to create an improved product.

[00:27:09] Speaker 03:
We have the same thing here, Your Honor, where we have this human engineered parameter of at least a 50% conversion rate that we can't achieve by growing plants and extracting red X from the plants.

[00:27:21] Speaker 03:
We can only achieve that by going to the lab.

[00:27:24] Speaker 03:
And that's what these inventors

[00:27:27] Speaker 03:
discovered that they gave rise to this specific application where you take these starting materials, some natural, they could be recombinant, you go to the lab and you contact them and you're able to create a hyperproduction of Rebex.

[00:27:43] Speaker 02:
There's nothing here that tells you how to do that.

[00:27:46] Speaker 03:
Your Honor, the specification

[00:27:49] Speaker 03:
does give working examples for the specific types of materials and the specific proportions that you would use in the lab to create certain things, catalyze the reaction.

[00:28:00] Speaker 03:
And that's the role of the specification to disclose and enable.

[00:28:05] Speaker 03:
It's not required that the claims spell out every technical step in that process to survive a section 101 challenge, rather the fact that they're claiming nonnaturally

[00:28:18] Speaker 03:
occurring conversion rates is enough to say... The claims are not limited to the examples, right?

[00:28:23] Speaker 02:
Correct, Your Honor, not limited to the examples.

[00:28:26] Speaker 02:
Okay, I think we're out of time.

[00:28:27] Speaker 02:
Let's have any questions.

[00:28:30] Speaker 02:
Thank you, Your Honor.

[00:28:34] Speaker 02:
Mr. Rosendahl?

[00:28:41] Speaker 00:
May it please the Court?

[00:28:42] Speaker 00:
I propose to start with the written description issue and I would like to

[00:28:48] Speaker 00:
sort of point to what I think are the undisputed facts on what was and wasn't known about this class of enzymes in the prior art.

[00:28:57] Speaker 00:
And I think perhaps the clearest place, the clearest sort of tight summary of what was known is at appendix pages 7507 to 08, which is from Pure Circle's own statement of undisputed facts about what was known.

[00:29:16] Speaker 00:
And this is pointing to some of the same material that my friend pointed to at other portions of the appendix.

[00:29:23] Speaker 00:
This is sort of the summary of it.

[00:29:43] Speaker 00:
And so the point is that the 12 known or suspected... It takes a while to fish out the appendices.

[00:29:58] Speaker 00:
It is a multi-volume extravaganza we have here.

[00:30:02] Speaker 02:
My apologies.

[00:30:04] Speaker 02:
It's not a deficiency in your appendix, it's a deficiency in my looking at the wrong appendix.

[00:30:10] Speaker 02:
Go ahead.

[00:30:11] Speaker 00:
So if we look at 7508, we get to the 12 known or suspected glucosal transferases.

[00:30:21] Speaker 00:
And those are found in the Richmond article.

[00:30:24] Speaker 00:
And the Richmond article itself is in the Joint Appendix, starting at page 5972.

[00:30:29] Speaker 00:
And what Richmond did was to take a database of expressed sequence tags from the Stevia rebaudiana Bertoni plant.

[00:30:40] Speaker 00:
express sequence tags are little snippets of DNA that have been derived from messenger RNA that's encoding proteins in the plant.

[00:30:48] Speaker 00:
And so we know that these are bits of DNA that are associated with some sort of protein that can be found in the leaf.

[00:30:54] Speaker 00:
And what he did was he searched the database using what he said was a sequence that identifies UDP glycosyl transferases, which is to say

[00:31:06] Speaker 00:
enzymes that can take a uridine diphosphate molecule with some sort of carbohydrate attached to it, not just glucose, and transfer it to something.

[00:31:18] Speaker 00:
It could be anything.

[00:31:18] Speaker 00:
It does not have to be a steveoglycoside.

[00:31:21] Speaker 00:
It does not have to be any sort of particular substrate.

[00:31:25] Speaker 00:
And so we have this immense category of enzymes, the so-called superfamily of UGT or

[00:31:34] Speaker 00:
glycosyl transferases, which are just the collection of anything that can transfer any sugar or carbohydrate from UDP to anything else.

[00:31:46] Speaker 00:
What about the testimony that he's relying on?

[00:31:53] Speaker 02:
about common structure.

[00:31:56] Speaker 02:
Because, of course, for written description purposes, if there is common structure, even though we're talking about a large number of possible candidates, that's sufficient for written description purposes.

[00:32:08] Speaker 02:
And he pointed us to some testimony, which he says that there's common structure.

[00:32:13] Speaker 02:
I guess it would have to be common structure in the active regions.

[00:32:17] Speaker 02:
And that that satisfies written description.

[00:32:19] Speaker 02:
That seems, in fact, to be the primary argument.

[00:32:23] Speaker 00:
Right.

[00:32:23] Speaker 00:
And our point, and I think the reason why that argument doesn't work will be clearer if I go through where I was heading before with Richmond and how we got to the 12 or the 5 enzymes.

[00:32:35] Speaker 00:
But the answer is there's no evidence of what that structure is.

[00:32:38] Speaker 00:
We can search in the specification.

[00:32:40] Speaker 00:
We can search in the prior bar.

[00:32:42] Speaker 00:
We can search in the record of this case.

[00:32:44] Speaker 00:
and there won't be any statement of what it is.

[00:32:48] Speaker 01:
The volunteer clearly says there is common structure in the right place but doesn't tell us what it is.

[00:32:53] Speaker 00:
Correct.

[00:32:54] Speaker 00:
He doesn't tell us because what we have to sort of cut to the chase is we have one class of enzymes, the UGT76G1 enzymes, and it's not just one enzyme, it's a whole family of enzymes.

[00:33:09] Speaker 00:
that have been shown to be able to make REV-M.

[00:33:13] Speaker 00:
And the only specific example that we know of that actually makes REV-M is the one that is the wild-type found in stevia rubidiana-britone, and we have an amino acid sequence for that one.

[00:33:25] Speaker 00:
There is no other enzyme that has been shown to make REV-M.

[00:33:31] Speaker 00:
And so the question is, what is it about this enzyme that allows it to catalyze?

[00:33:36] Speaker 00:
What is it that other enzymes would have to have in order to be able to carry out this function?

[00:33:42] Speaker 00:
And there isn't any answer to that question.

[00:33:44] Speaker 02:
I'm a little confused.

[00:33:45] Speaker 02:
I thought that there was agreement that there were four or maybe five enzymes that made the conversion.

[00:33:52] Speaker 00:
No, Your Honor.

[00:33:52] Speaker 00:
And that's what I was trying to get to with my Richmond example.

[00:33:55] Speaker 00:
So if I could just pick up there.

[00:33:57] Speaker 00:
So what Richmond did was he took a sequence

[00:34:00] Speaker 00:
that has been identified as being associated with the part of these enzymes that grabs onto the uridine diphosphate molecule.

[00:34:08] Speaker 00:
And that is the so-called signature sequence for what they say is the signature sequence for glucosal transferases.

[00:34:15] Speaker 00:
In fact, it's the signature sequence for glycosal transferases, and it's the signature for what grabs the uridine molecule.

[00:34:22] Speaker 00:
It doesn't tell us anything about what substrate the enzyme works on.

[00:34:27] Speaker 00:
And this is all spelled out.

[00:34:28] Speaker 00:
in the joint appendix.

[00:34:29] Speaker 00:
So what he does is he uses that, he sticks that into the database and says, okay, what in stevia rebaudiana in this database of expressed sequence tags has this sequence that's a candidate even to be in this super family of enzymes?

[00:34:43] Speaker 00:
And he finds 12.

[00:34:45] Speaker 00:
And of the 12, seven of them don't do any transfers that they could find.

[00:34:50] Speaker 00:
And in fact, he got 54 hits in the database.

[00:34:54] Speaker 00:
There was only a complete enzyme

[00:34:57] Speaker 00:
sequence for 12 of them.

[00:34:59] Speaker 00:
And when he tested them on various substrates, seven of them didn't do anything.

[00:35:04] Speaker 00:
Five of them were able to do some sort of glucose transfer, but only three of them were able to transfer to steval glycoside substrates.

[00:35:13] Speaker 00:
And that is what is summarized here on page 7508.

[00:35:16] Speaker 00:
He says, only five were shown to be functional, only three were capable of adding glucose units to steval glycosides.

[00:35:25] Speaker 00:
And then he says, later work showed that another enzyme also is able to at least transfer sugars to stebial glycosides.

[00:35:33] Speaker 00:
But none of these enzymes has been shown to be able to make RebM, except for the one working example that is described in the specification here.

[00:35:46] Speaker 02:
And so, you know, this gets you... So you say that.

[00:35:50] Speaker 02:
Where do we find in the record testimony that says that these other enzymes are not able to make RebM?

[00:36:01] Speaker 00:
Well, I think there's just the absence of any indication that any of them has been able to make RebM except for the one that we're talking about.

[00:36:12] Speaker 00:
I mean, I don't know how you would prove the negative.

[00:36:15] Speaker 02:
Well, I mean, you could have testimony that these other enzymes haven't been shown to make Reb M. Well, I guess what we know about the other enzymes in the prior ARC... I mean, you've got to understand.

[00:36:30] Speaker 02:
It can be very articulate.

[00:36:32] Speaker 02:
It can be very knowledgeable.

[00:36:33] Speaker 02:
That's not the record.

[00:36:34] Speaker 02:
The record is what we have to deal with.

[00:36:36] Speaker 02:
What I want is something in the record that tells me that what you're saying is correct.

[00:36:44] Speaker 00:
Your Honor, I guess I don't think that that issue is actually in dispute.

[00:36:53] Speaker 00:
I don't think there's ever been a suggestion that any enzyme other than the one example mentioned in the specification can turn REV-D into REV-M.

[00:37:03] Speaker 00:
That's part of the reason why we have a breadth problem here.

[00:37:07] Speaker 00:
First of all, we don't even know for the five

[00:37:10] Speaker 02:
Is there nothing in the record that says that these other enzymes aren't capable of making Rev D into Rev M?

[00:37:21] Speaker 00:
Your Honor, all I can tell you is that there's no evidence of any of them doing it.

[00:37:27] Speaker 00:
And our point is that that's one reason why we have a problem knowing which enzymes are going to be in this genus and which ones are not.

[00:37:40] Speaker 00:
it's up to the specification, if they're going to claim every possible enzyme that can make RebM, they have to either show some structural feature, some structure-function relationship that explains why it is that the enzyme can do this, or else they need to come up with a representative number of species, and they haven't even argued that their one example is a representative number of species.

[00:38:06] Speaker 00:
So we have just the one, and instead of

[00:38:10] Speaker 00:
saying, well, we're going to try to make a bunch more, see if we can find any more, and then try to get a representative number of species.

[00:38:17] Speaker 00:
They just say, well, we have the one, and with homology modeling, you could probably find some others.

[00:38:22] Speaker 00:
And it's up to you to sort of go out and find them.

[00:38:26] Speaker 00:
And I think under this report's cases, that is

[00:38:28] Speaker 00:
I'm not enough to support the written description requirement.

[00:38:32] Speaker 01:
But in our procedural posture, it's up to you to show that no reasonable juror could review this record taking in the light most favorable to them and reach any conclusion other than that you've proven by clear and convincing evidence a lack of written description.

[00:38:48] Speaker 01:
Is that correct?

[00:38:49] Speaker 00:
Well, I think that's a fair statement of the law, Your Honor.

[00:38:51] Speaker 00:
But what we have undisputed facts on is that there wasn't any

[00:38:56] Speaker 00:
structure, sort of three-dimensional structure known for these enzymes as of 2012.

[00:39:02] Speaker 00:
All that was known were amino acid sequences.

[00:39:05] Speaker 00:
And we also have no dispute that amino acid sequences are an extremely poor guide to enzyme functionality.

[00:39:11] Speaker 02:
What about the testimony that he's cited to us about common structural features?

[00:39:16] Speaker 00:
Right, but the common structural features that he's pointed to are two, your honor.

[00:39:20] Speaker 00:
One of them is this supposed signature amino acid sequence, which is just the

[00:39:26] Speaker 00:
super family's sequence that grabs onto the uridine molecule.

[00:39:30] Speaker 00:
So that doesn't tell you even what operates on steveal glycosides, let alone what can make RevM.

[00:39:35] Speaker 00:
And the other thing is he points to this computer model of a three-dimensional structure of this one enzyme, which first of all was made in 2016, long after the patent application was filed.

[00:39:50] Speaker 00:
and which we don't know if it's correct or not.

[00:39:54] Speaker 00:
But even if it is correct, they haven't told us what parts of that structure are the ones that account for the functionality.

[00:40:01] Speaker 00:
And it's just the one molecule.

[00:40:03] Speaker 00:
So again, we don't know.

[00:40:04] Speaker 01:
Why doesn't that translate into a genuine disputed material fact, where a jury could say at the end of all that, you've not proven by clear and convincing evidence that it wouldn't be?

[00:40:14] Speaker 01:
what they're saying it would do, or that a person of skill in the art wouldn't understand what Dr. Bollinger says they would understand.

[00:40:21] Speaker 00:
No, but the argument that's being made, Your Honor, is that every enzyme has to fit its substrate, right?

[00:40:27] Speaker 00:
And it's true enough that enzymes, in order to work, are presumed to fit their substrates.

[00:40:32] Speaker 00:
But the argument made at that level of generality, which is all that Dr. Bollinger did, is not enough

[00:40:41] Speaker 00:
to identify a common structural feature.

[00:40:46] Speaker 01:
The patent office looked at all this as well, including the homology question.

[00:40:50] Speaker 01:
Did they know it?

[00:40:51] Speaker 00:
Well, the patent office, I guess, considered in the scope of its prior decision, did address the

[00:41:09] Speaker 00:
written description question and focus on... Have you looked at homology modeling?

[00:41:12] Speaker 01:
Said it was advanced?

[00:41:13] Speaker 01:
Said it could help screen the mutants?

[00:41:15] Speaker 00:
Well, no, that whole discussion of homology modeling was all in the enablement portion of the decision.

[00:41:20] Speaker 01:
I think what the... It has no implications for a genuine dispute of material fact over written descriptions?

[00:41:26] Speaker 00:
I don't believe so, Your Honor.

[00:41:27] Speaker 01:
I think that this Court has never found that because homology... Was it in the record that the District Court could consider

[00:41:35] Speaker 01:
in determining whether there was a material dispute of fact on written description.

[00:41:40] Speaker 00:
I'm sorry, was what in the record, Your Honor?

[00:41:42] Speaker 01:
The PTO's analysis.

[00:41:45] Speaker 00:
I suppose the PTO's final decision was submitted to the court in connection with the judgment.

[00:41:52] Speaker 01:
Why doesn't that contribute to a genuine dispute of material fact that a reasonable juror, through Dr. Bollinger, might say is relevant to written description?

[00:42:00] Speaker 00:
Well, I'm not sure that a reasonable juror would be permitted to consider that the PTAB's final written decision.

[00:42:06] Speaker 01:
Not excluded from the record at summary judgment stage yet.

[00:42:08] Speaker 00:
Well, but in any event, I think what the district court had to decide was on the evidence before it whether there was a genuine issue of dispute of material fact.

[00:42:19] Speaker 00:
And what the PTAB decided on.

[00:42:23] Speaker 02:
I think we have cases that say that the PTAB decision is not something that can be presented

[00:42:29] Speaker 00:
I think that's right.

[00:42:30] Speaker 00:
I've certainly never been in a number of jury cases, and I've never had a judge permit the PTAB decision to be shared with the jury.

[00:42:36] Speaker 00:
But be that as it may, I think the real point is that the PTAB made, I think, an erroneous decision based on whatever record was before it.

[00:42:46] Speaker 00:
What it did not have before it was the stipulated claim construction in this case.

[00:42:50] Speaker 00:
It did not have Dr. Bollinger's testimony that, for example,

[00:42:57] Speaker 00:
And any enzyme that carries out the function, no matter what the structure is, would fall within the scope of the claims.

[00:43:04] Speaker 00:
And you'll find that at appendix 61, 15 to 16.

[00:43:08] Speaker 00:
And similar statements by Dr. Bollinger.

[00:43:12] Speaker 02:
Where is that last one, 61 what?

[00:43:15] Speaker 00:
61, 15 to 16.

[00:43:16] Speaker 00:
He was asked.

[00:43:36] Speaker 01:
That's your main admission that you point to, that long, hypothetical, various assumptions, and at the end of it he says yes.

[00:43:46] Speaker 00:
Well, the question is, there's an enzyme that doesn't belong to the structural family that was known as the filing of the patent, but it still performs the function listed in the court's claim construction.

[00:43:57] Speaker 00:
And then the answer is, yeah, if it performs the function regardless of the structure,

[00:44:04] Speaker 00:
it falls within the scope of the claim.

[00:44:06] Speaker 00:
He could have said, oh no, you don't understand.

[00:44:08] Speaker 00:
A UDP is a very known thing.

[00:44:10] Speaker 00:
If it didn't have the structure that we all know about, then it wouldn't be within the scope of the claim.

[00:44:14] Speaker 00:
But that's not what he said.

[00:44:15] Speaker 00:
He said anything that carries out the function is going to fall within the scope of the claim.

[00:44:20] Speaker 00:
So I think that shows that this is a functional definition, and that is not something that the PTAB had before it.

[00:44:25] Speaker 01:
And was that definition one that was mandated by the specification, in your view, or I know it was done by stipulation, but

[00:44:34] Speaker 00:
So there is a similar statement in the specification.

[00:44:39] Speaker 00:
The court can find it, for example, at appendix 127 in column 2, starting at about line 30.

[00:44:45] Speaker 00:
And it says the UDP glucose transferase can be any UDP glucose transferase capable of adding at least one glucose unit, and then the rest of it is the same.

[00:44:58] Speaker 00:
I think the argument that's being presented here is that, well, the specification talks about UDP-glucosal transferases, and when we, PureCircle, proposed that, we took out the word UDP-glucosal transferase and substituted it with a type of enzyme that can.

[00:45:16] Speaker 00:
We think that changing what the spec said to what they said actually broadens the class to any kind of enzyme regardless of structure.

[00:45:27] Speaker 00:
even if the court were not to read it that way, we would still have a functionally defined genus.

[00:45:34] Speaker 01:
What about Claim 14?

[00:45:35] Speaker 01:
Why is there lack of adequate written description for that narrower claim?

[00:45:39] Speaker 00:
So Claim 14, again, is not just a single enzyme.

[00:45:46] Speaker 00:
We have an admission from Dr. Bollinger, their expert, at Appendix 6363.

[00:45:53] Speaker 00:
that the UGT76G1 subclass, the members of that class vary considerably in amino acid sequence and properties.

[00:46:01] Speaker 00:
And so again, the record is not super clear on the taxonomy of this superfamily of UGT enzymes, but if one Googles it, as I did, one will see that they are

[00:46:16] Speaker 00:
biogenetically and then structurally classified.

[00:46:19] Speaker 00:
So the numbers in these enzyme names correspond to different classes of plants or animals.

[00:46:25] Speaker 00:
So 1 to 50 is animals, and 51 to 70 is yeasts and fungi, and 71 to 100 is plants.

[00:46:33] Speaker 00:
And so in plant family 76, whatever that is, which contains a bunch of different plants, somewhere in that family was the first one of these that was found.

[00:46:42] Speaker 00:
And so it was named

[00:46:43] Speaker 00:
because it was found in plant family 76, it was called 76, and then there are other categories.

[00:46:50] Speaker 00:
So when you talk about 76G1, you're not talking about a single enzyme.

[00:46:54] Speaker 00:
You're talking about a family of enzymes that could be found in a bunch of different plants that are variable and have different properties.

[00:47:01] Speaker 00:
And so it's just not true that there's a specific answer in that claim.

[00:47:08] Speaker 01:
They outline a gray brief from page six atop of page seven, several genuine disputes of material fact that they say should that minimum have concluded summary judgment.

[00:47:22] Speaker 01:
Just if you can give me, I don't want you to go through like six sentences, but I take it your view is going to be none of these are genuine disputes of material fact, and I don't know if there's another general response you would have to that.

[00:47:36] Speaker 00:
Well, I think that some of them are not material.

[00:47:40] Speaker 00:
I mean, for example, the question whether they did or didn't need the broad genus definition of proven infringement, I don't think really is going to move the needle on validity.

[00:47:48] Speaker 00:
So that's not a material dispute.

[00:47:50] Speaker 00:
But I don't think there is any genuine dispute of material fact about what was known about this class of enzymes structurally at the time.

[00:47:59] Speaker 00:
It was absolutely clear that in 2012, the three-dimensional structure was not known and that the

[00:48:06] Speaker 00:
amino acid sequences were a poor guide to the function of the enzymes.

[00:48:11] Speaker 00:
And then the structure-function correlation, again, there is no... I think the key fact that was not brought out previously is that everyone agrees, and there's no dispute,

[00:48:25] Speaker 00:
you can't tell either from the homology model picture or from the amino acid sequence whether any given enzyme will work to make RebM, you have to test them all.

[00:48:39] Speaker 00:
And so the fact that you have to test each and every one of them to find out whether they do or don't fit within the genus under this court's cases

[00:48:48] Speaker 00:
is not enough to satisfy the written description.

[00:48:50] Speaker 02:
I'm not sure that's true if you're talking about a limited number that have to be tested.

[00:48:56] Speaker 02:
And so perhaps this turns on how many potential candidates there are.

[00:49:03] Speaker 02:
And my understanding is that they

[00:49:09] Speaker 02:
they say that there are common structural features that limit the number of enzymes that would have to be tested.

[00:49:17] Speaker 02:
And I'm looking again at 6115.

[00:49:22] Speaker 02:
What is the question here with an enzyme that doesn't belong to the structural family?

[00:49:27] Speaker 02:
What's the structural family?

[00:49:30] Speaker 00:
Well, again, this was done in the context of this argument that UDPs have

[00:49:36] Speaker 00:
have a classification that can be named based on their amino acid sequence.

[00:49:41] Speaker 00:
And the argument was being advanced by Dr. Bollinger that when you use the word UDP, you are referring to something that has a structure.

[00:49:52] Speaker 00:
And the question was posed, okay, well, if you have something that doesn't have that structure, that doesn't belong in that family of structures that you've been talking about, would that still infringe?

[00:50:01] Speaker 00:
And he said, yes, it would.

[00:50:02] Speaker 02:
So the structure that's being referred to here is a structure defined by amino acid sequences?

[00:50:08] Speaker 00:
Well, again, we know that you can name enzymes based on their amino acid sequences.

[00:50:14] Speaker 00:
So they're named based on how closely they match up with

[00:50:19] Speaker 00:
other amino acids that have been previously found.

[00:50:22] Speaker 02:
I'm sorry to be so persistent about this, but it's an important question.

[00:50:28] Speaker 02:
In the opening argument, we talked a lot about whether there's a common structure here.

[00:50:33] Speaker 02:
He identified some testimony from their expert witness that says there are common structural features.

[00:50:41] Speaker 02:
And I'm trying to understand whether, first of all, there are common structural features, and if so, what they are.

[00:50:49] Speaker 00:
Okay.

[00:50:52] Speaker 00:
The record doesn't reflect any structural features common to enzymes that can make RebM.

[00:51:03] Speaker 00:
What we have in terms of common structural features is a sequence that has been associated with the superfamily of enzymes of glycosyl transferases

[00:51:16] Speaker 00:
that can take sugar molecules from UDP.

[00:51:19] Speaker 00:
So we do have that.

[00:51:20] Speaker 00:
But that doesn't answer the question, what can we?

[00:51:23] Speaker 02:
So the answer is that the common structural features that have been identified are common structural features that identify a class of compounds that could be tested and might produce revm.

[00:51:34] Speaker 00:
Yes, your honor.

[00:51:38] Speaker 00:
To the extent that they've been identified at all.

[00:51:40] Speaker 00:
And again, the three dimensional features that we're talking about have not been identified at all.

[00:51:45] Speaker 00:
They say, well, you can look at this thing from 2016, four years after the patent, and you could kind of figure them out.

[00:51:50] Speaker 00:
But we don't know what they are.

[00:51:51] Speaker 00:
And there hasn't been any correlation between any amino acid sequence or any three-dimensional structure and the effectiveness in catalyzing the reaction.

[00:52:02] Speaker 00:
I see that I'm over my time.

[00:52:03] Speaker 00:
I would like to spend just a moment on the 101.

[00:52:07] Speaker 00:
So I think that the key point here

[00:52:11] Speaker 00:
on 101 is that we have only one working example in the specification, and it replicates precisely the process that is carried out in the plant.

[00:52:22] Speaker 00:
It starts with a naturally occurring RebD, it uses an enzyme that is structurally and functionally identical to the enzyme found in the plant, and it produces RebM, which is also a naturally occurring compound.

[00:52:41] Speaker 00:
there isn't any doubt that the claims are directed to a naturally occurring phenomenon, and I think the key point here is that everyone agrees that the claims are broad enough to encompass enzymes that are structurally and functionally identical to the naturally occurring enzymes, which is not surprising because the only example they have of one that works comes from the plant.

[00:53:04] Speaker 00:
And that's why we think that it's

[00:53:06] Speaker 00:
clear under ALSTEP 1 that these are directed to natural phenomenon, and they haven't even made an argument at ALSTEP 2 that purification or percent conversion are anything other than routine and conventional laboratory steps.

[00:53:20] Speaker 01:
I recognize you take the view that method of preparation claims are not per se eligible, but is it relevant to the analysis that they contend it's a method of preparation claim?

[00:53:30] Speaker 00:
Well, there's certainly no categorical immunity for methods of preparation or manufacture.

[00:53:35] Speaker 00:
I think American Axel would be an example of a method of manufacture that ran into a 101 problem.

[00:53:39] Speaker 00:
I think the other thing to bear in mind is that if you look at the method of preparation cases that were successful in avoiding

[00:53:47] Speaker 00:
101 problems like Illumina and CellsDirect, the claims themselves had steps in them that were about laboratory methods that were distinct from the natural principle that was being... It was more than a result.

[00:54:05] Speaker 00:
It was more than a result, right.

[00:54:07] Speaker 00:
It was more than a result.

[00:54:07] Speaker 00:
It was a method.

[00:54:08] Speaker 00:
It had concrete steps to achieve the result, and that was what distinguished those cases from this one, where here

[00:54:15] Speaker 00:
We just have a wish for a 50% conversion or a wish for a particular level of purity and we don't have anything beyond just put the substrate in with the enzyme and wait until you have enough.

[00:54:30] Speaker ?:
Okay.

[00:54:31] Speaker ?:
Thank you.

[00:54:43] Speaker 02:
So, just first address would you this question.

[00:54:47] Speaker 02:
The common structural features only identify potential candidates for conversion to RebM.

[00:54:57] Speaker 02:
They don't identify those that convert to RebM.

[00:55:01] Speaker 03:
Correct, Your Honor.

[00:55:03] Speaker 03:
There is automated assay testing.

[00:55:05] Speaker 03:
that both parties' experts agreed was well understood and routine at that time that you would then use to validate.

[00:55:12] Speaker 03:
And that's why, Your Honor, the

[00:55:15] Speaker 03:
structure-function correlation test in Juneau is also very instructive here, where the course said it doesn't need to be a perfect correlation, but there's very much a dispute about whether the tools a skilled artisan would have would be adequate here, because what the art showed is that you use homology modeling to narrow down what has a common need.

[00:55:36] Speaker 02:
The question is, how many candidates are there?

[00:55:39] Speaker 02:
The 1824 is not the correct number, because that's only one enzyme.

[00:55:43] Speaker 02:
So you say you'd have to at least multiply that by four because of the other enzymes that have been identified as having possibilities, right?

[00:55:58] Speaker 02:
Correct, Your Honor.

[00:56:00] Speaker 02:
But just the next step, so that if there are other enzymes, it would be even more than the 9,000 that would be presented by that calculation, right?

[00:56:11] Speaker 02:
No, Your Honor, not that we're known at the time.

[00:56:14] Speaker 02:
No, but we've established that this claim covers unknown enzymes, right?

[00:56:21] Speaker 02:
Correct.

[00:56:21] Speaker 02:
So is that possible?

[00:56:23] Speaker 03:
Yes, Your Honor, which is always possible in science, and yet that hasn't been a basis on which this Court has invalidated claims.

[00:56:31] Speaker 02:
And Dr. Bollinger did... Yeah, but has this Court said that we don't look at possible additional candidates if the claim is that broad?

[00:56:41] Speaker 03:
I'm not aware of a case, Your Honor, that addresses it in those terms, but the four area factors are existing knowledge, extending content of the prior arts.

[00:56:50] Speaker 03:
You look at what is known, and then you have maturity of science or technology and predictability of the aspect at issue.

[00:56:57] Speaker 02:
which would give you some sense of how much is known and how much might remain to be explored.

[00:57:14] Speaker 03:
correct your honor, but you assess whether written description is met, as a matter of fact, from the perspective of a skilled artisan at the time of filing.

[00:57:23] Speaker 03:
And a skilled artisan would have known about the UDP glucose transferases

[00:57:29] Speaker 03:
disclosing Richmond, discussing Keyshore, Keyshore use, homology modeling, identified specific amino acid substitutions that would work on these enzymes for mutations.

[00:57:41] Speaker 03:
And you have testimony from Dr. Bollinger, for example, at page 6119 of the appendix, where following on from the language that my colleague was quoting, what Dr. Bollinger says is that

[00:57:56] Speaker 03:
He believes that the a type of language is defining UDP commercial transfers according to what was known at the time of filing.

[00:58:03] Speaker 03:
And at the time, it associated everything into a known structural family.

[00:58:08] Speaker 03:
And Dr. Bollinger didn't know of any other structures that would fit into that, nor did Swigin as the party with the burden

[00:58:18] Speaker 03:
produce any evidence that would suggest, no, this is actually not mature and not predictable because a skilled artisan would have strong reason to believe that there is a completely different structural family that has an amino acid sequence that looks nothing like Richmond's signature sequence from 2005.

[00:58:37] Speaker 03:
looks nothing like the three-dimensional models produced with computer tools available in 2012.

[00:58:45] Speaker 03:
There's no evidence like that from Sweden, and therefore these questions about whether all of those tools, including the assay testing,

[00:58:53] Speaker 03:
would be part of what informs how a skilled artist would view the specification for the adequacy of disclosure.

[00:59:01] Speaker 03:
That's all very much disputed.

[00:59:11] Speaker 03:
point back again to claim 14 and the stipulated construction, which is a UGT enzyme encoded by a gene classified in UGT family number 76, subfamily G, and gene number 1 at page 5160 of the appendix, and therefore

[00:59:28] Speaker 03:
So even beyond all of the other arguments we've been discussing here, that itself is a purely structural definition that meets the original claims doctrine.

[00:59:39] Speaker 03:
And then even if you go on to look at the adequacy of written descriptions for mutants under the three tests in Juneau, we have these tools we've been discussing.

[00:59:49] Speaker 03:
Each party has their own view of whether those tools would

[00:59:54] Speaker 03:
inform adequate written description, and that dispute, Your Honors, should be resolved by a jury.

[00:59:58] Speaker 03:
Thank you.

[01:00:00] Speaker 02:
Thank you.

[01:00:00] Speaker 02:
Thank both counsels.

[01:00:01] Speaker 02:
The case is submitted.

[01:00:02] Speaker 02:
That concludes our section for this morning.