8339 SE 57th St., Mercer Island, WA 98040
`
`
`Via email
`05 June 2015
`
`Amy McCourt, Esq.
`Patent Attorney
`Roland H. Schwillinski, Esq.
`Sr. Director, IP and Commercial Litigation
`Illumina Inc.
`5200 Illumina Way
`San Diego, CA 92122
`
`Dear Ms. McCourt and Mr. Schwillinski:
`
`In your 13 April 2015 letter, you asked for “an explanation” describing the question of
`infringement of The Scripps Research Institute patents and suggested a more specific
`identification of the relevant claims (by claim chart) and how “Illumina’s processes or
`products meet the requirements of the specific claim language.” We are happy to
`respond to your request in a cooperative manner and are providing the detailed
`explanation you requested. Scripps would much prefer to spend its time working on
`other pioneering inventions and medical research than engage in protracted and
`expensive litigation on an issue that should be readily resolvable. Let us begin by
`focusing on U.S. Patent 6,060,596 (‘596 Patent) that provides for the A-B-C formula.
`Illumina’s Sentrix and BeadArray products have an oligonucleotide structure attached to
`the beads as demonstrated by the following Illumina publications:
`
`1. Steinberg-Tatman et al. BIOCONJUGATE CHEMISTRY 17:841-848, (2006)
`(“Steinberg 2006”).
`2. Gunderson et al. GENOME RESEARCH 14:870-877 (2004) (“Gunderson 2004”).
`3. Gunderson et al. NATURE GENETICS 37:549-554 (2005) (“Gunderson 2005”).
`4. Steemers and Gunderson, Illumina Company Profile, Pharmacogenomics
`6(7):777-782, 2005 (“Illumina Profile”).
`5. Illumina Technical Bulletin entitled “Illumina Gene Expression Profiling” © 2005
`Illumina, Inc. (“Tech Bulletin 2005”).
`
`We can provide pdf files of the references if you do not have them.
`
`The Illumina Profile provides a 2005 description of the oligonucleotide structure on each
`bead and gives an overview of the BeadArray and Sentrix manufacturing process by
`stating: “This decoding process simultaneously provides a map identifying each bead,
`along with a measure of their functionality. [reference: Gunderson 2004].” The Tech
`Bulletin 2005 on the second page under the section “Bead Design” describes that
`“Oligonucleotide probes attached to beads in Illumina’s Focused Assays are
`concatamers of a 29-base IllumiCode address and a 50-base gene-specific probe.
`The IllumiCode is used to decode the array, the gene-specific probe is used for
`analysis of expressed transcripts (Figure 3).” (emphasis added). Therefore, chemical
`moiety A is the gene-specific probe and the decoding sequence was given a name in
`
`
`
`1
`
`
`
`Page 1 of 10
`
`ILMN EXHIBIT 1046
`
`

`
`
`
`2005, the “IllumiCode” which was called the “identifier oligonucleotide C” in the ’596
`Patent. According to the ’596 Patent specification, the linker molecule “ . . . can be any
`molecule that performs the function of operatively linking the chemical moiety to
`identifier oligonucleotide.” See col. 8, ll. 20-24. For example, the linker (-B-) in the
`Illumina Profile is the phosphate group that operatively links the gene-specific probe and
`the decoding sequence together.
`
`The Gunderson 2005 publication at page 550 describes the Infinium assay, whereby
`50-base capture probes (chemical moiety A) and 25-base decoding sequences
`(oligonucleotide C) are linked (Linker B) as a single oligonucleotide of 75 bases. The
`Illumina Profile describes at page 780 that the Infinium assay was designed to genotype
`100,000 SNPs per sample. The Gunderson 2004 publication describes Illumina’s
`decoding process and that the bottleneck of decoding 100,000s of bead types could be
`processed using pairs of decoding sequences at page 874. Two 25-base decoding
`sequences linked to one 50-base gene-specific probe sequence would result in one
`very long (100-mer) oligonucleotide (Group C in Steinberg 2006).
`
`The Steinberg 2006 publication describes how Illumina’s bead types, which makes up
`the BeadArray and Sentrix products, are made. Steinberg 2006 describes several
`schemes (i.e., Groups A, B, and, C) for the attachment of two different oligonucleotides
`and arrives at the best solution, resulting in an A-B-C bifunctional molecule.
`
`
`“Group C schemes, while this study was carried out, involving one very
`long (100-mer) oligonucleotide, were not considered because low
`synthesis yields would be expected. Group A type schemes were
`evaluated first. . . . Group A schemes, however, have an inherent
`problem. . . . In schemes from group B the aforementioned problem
`[difficultly to control the ratio of immobilized oligonucleotides] is avoided
`as only one orthogonally protected amino acid is coupled to the
`amino beads.” Steinberg 2006 at p. 844 (emphases added).
`
`Here is a claim chart showing how the Illumina BeadArray and Sentrix products infringe
`Claims 1, 3, 10, and 16 of U.S. Patent 6,060,596 (’596 Patent).
`
`
`
`
`2
`
`
`
`Page 2 of 10
`
`

`
`
`
`Claim
`
`
`1.
`
`Claim Language
`
`[1] A bifunctional molecule
`according to the formula A-B-C,
`
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`The BeadArray product uses a bifunctional
`molecule according to the formula A-B-C,
`where A and C are two different
`oligonucleotides. As referenced above, B
`can be any molecule and in this instance is
`an orthogonally protected lysine linker
`(“Group B” in Steinberg 2006)1.
`
`“We developed a simple and elegant
`synthesis strategy that enables us to attach
`controlled (equimolar) amounts of two
`different oligonucleotides onto one silica
`bead.” Steinberg 2006 at abstract
`(emphasis added).
`
`“Figure 1 illustrates three potential schemes
`for attaching two different
`oligonucleotides to the same bead. In
`this paper, we focus on methods for
`groups A and B.” Id. at p. 841 (emphases
`added).
`
`
`Figure 1 at p. 841.
`
`“To enable sequential attachment of two
`distinct moieties to one bead, we
`developed an alternative chemistry scheme
`involving the use of orthogonally
`protected lysine. The two amino groups
`on the lysine are sequentially activated,
`permitting sequential addition of two
`
`
`1
`
` In addition to col. 8, ll. 20-24 of the ’596 specification, “linker molecule” is also identified as “-B-.” See
`col. 8, ll. 35-38 (“A linker molecule can vary in structure and length, and provide at least two features: (1)
`operative linkage to chemical moiety A and (2) operative linkage to identifier oligonucleotide C.”). This
`means that even Group C in Steinberg 2006 (presumably arranged as Address1–Address2–Gene
`Probe), the decoding sequence–capture probe sequence in Gunderson 2005, and the “IllumiCode Gene-
`specific Probe” in Tech Bulletin 2005 all have a phosphate group that is representative of the linker
`molecule (-B-) that operatively links the chemical moiety (A) and the identifier oligonucleotide (C) as a
`single oligonucleotide attached to the bead.
`
`
`
`3
`
`
`
`Page 3 of 10
`
`

`
`
`
`Claim
`
`
`
`
`Claim Language
`
`[2] wherein A is a polymer
`comprising a linear series of
`chemical units represented by
`the formula (Xn)a, wherein X is a
`single chemical unit in polymer
`A,
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`different oligonucleotides.” Id. at p. 841
`(emphases added).
`The BeadArray product uses a polymer (A)
`comprising a linear series of chemical units
`represented by the formula (Xn)a, (e.g., “Xn”
`is a nucleotide such as A, C, G, or T and “a”
`is the length of the oligonucleotide, such as
`25 bases), wherein X is a single chemical
`unit in polymer A [i.e., oligonucleotide].
`
`“We developed a simple and elegant
`synthesis strategy that enables us to attach
`controlled (equimolar) amounts of two
`different oligonucleotides onto one silica
`bead.” Steinberg 2006 at abstract
`(emphasis added).
`
`“Figure 1 illustrates three potential schemes
`for attaching two different
`oligonucleotides to the same bead. In
`
`this paper, we focus on methods for
`groups A and B.” Id. at p. 841 (emphases
`added).
`
`Figure 1 at p. 841.
`
`Steinberg 2006 further illustrates
`oligonucleotide 2 Group B as the polymer
`[i.e., Assay seq]
`
`“[I]t is important to optimize the amount of
`
`
`
`4
`
`
`
`Page 4 of 10
`
`

`
`
`
`Claim
`
`
`Claim Language
`
`[3] B is a linker molecule
`operatively linked to A and C.
`[sic] and
`
`[4] identifier oligonucleotide C is
`
`
`
`
`
`
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`oligo that we need to immobilize on these
`double-attachment beads . . . in with
`address oligonucleotides (25 mers), beads
`with assay sequences (50mers), and
`beads with both oligonucleotides
`immobilized together on the two attachment
`beads (see Figure 6, model C).” Id. at p.
`846.
`
`Model C of Fig. 6 at p. 845.
`The BeadArray product uses a linker
`molecule (B) according to the ’596
`specification. The example in Steinberg
`2006 illustrates an orthogonally protected
`lysine that operatively linked to A and C.
`
`“To enable sequential attachment of two
`distinct moieties to one bead, we developed
`an alternative chemistry scheme involving
`the use of orthogonally protected lysine.
`The two amino groups on the lysine are
`sequentially activated, permitting
`sequential addition of two different
`oligonucleotides.” Id. at p. 841 (emphasis
`added).
`
`“Figure 4 shows the synthetic approach to
`sequential oligonucleotide attachment,
`as demonstrated in the potential scheme
`in Figure 1 group B.” Id. at p. 842
`(emphasis added, the red rectangle
`delineates the linker (Lys) moiety).
`
`Final product of Fig. 4 at p. 843, where the
`lysine linker (highlighted by the red box)
`links A [i.e., “Oligo 2”] and C [i.e., “Oligo 1”].
`The BeadArray product uses an identifier
`
`5
`
`
`
`Page 5 of 10
`
`

`
`
`
`Claim
`
`
`Claim Language
`
`represented by the formula
`(Zn)a,
`
`[5] wherein a unit identifier
`nucleotide sequence Z within
`oligonucleotide C identifies the
`chemical unit X at position n; and
`wherein n is a position identifier
`for both X in polymer A and Z in
`oligonucleotide C having the
`value of l+i where i is an integer
`from 0 to 10, such that when n is
`1, X or Z is located most
`proximal to the linker, and
`
`
`
`
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`oligonucleotide C is represented by the
`formula (Zn)a.
`
`“1st Sequence: S9 5’ (NH2)
`TTTGATGTCCCATTCCCCACGCGTT
`2nd Sequence: S13 5’ (NH2 or CHO)
`TTTGATCGTAGCCGGTATGCGACGG.”
`Id. at p. 846, Table 1. The “1st Sequence”
`meets the formula (Zn)a where Zn = A, C,
`G, or T and a = 25. The “2nd sequence”
`meets the formula (Zn)a where Zn = A, C,
`G, or T and a = 25.
`The BeadArray product uses an identifier
`nucleotide sequence to identify polymer A.
`For example, for both the 1st and 2nd
`sequences described in the last element,
`the first “T” nucleotide is located most
`proximal to the linker.
`
`“Model beads with two address
`sequences (Figure 6, model B) were
`synthesized, . . . Table 1 summarizes
`hybridization intensities for both the first
`
`and second address sequences, . . . ” Id. at
`p. 846 (emphases added).
`Model B of Fig. 6 at p. 845.
`
`“Random bead loading combined with
`decoding, described in detail previously (6).
`. . For very large numbers of bead types
`(tens to hundreds of thousands), a more
`practical decoding approach uses two
`different address oligos on each bead.”
`Id. at 841 (emphasis added).
`
`Steinberg 2006 cites reference 6
`[Gunderson et al. GENOME RESEARCH
`14:870-877 (2004) (“Gunderson 2004”) to
`describe the decoding process in detail.
`Gunderson 2004 describes how the
`
`6
`
`
`
`Page 6 of 10
`
`

`
`
`
`Claim
`
`
`Claim Language
`
`
`
`[6] a is an integer from 4 to 50.
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`identifier oligonucleotide C identifies the
`nucleotide sequence of polymer A.
`
`“Our approach uses longer sequences,
`each designed to hybridize to a defined
`target with high specificity. It is capable of
`decoding, with high accuracy, many
`1000s of bead types. Each bead type is
`defined by a unique DNA sequence that
`is recognized by a complementary
`decoder.” Gunderson 2004 at p. 870.
`
`“[I]t is important to optimize the amount of
`oligo that we need to immobilize on these
`double-attachment beads . . . in with
`address oligonucleotides (25mers),
`beads with assay sequences (50mers),
`and beads with both oligonucleotides
`immobilized together on the two attachment
`beads (see Figure 6, model C).” Steinberg
`2006 at p. 846.
`
`Model C of Fig. 6 at p. 845.
`The BeadArray product uses oligonucleotide
`sequences where a [i.e., the length of the
`oligonucleotide] is 29 nucleotides [i.e.,
`29mers] (Tech Bulletin 2005) and 50
`nucleotides [i.e., 50mers] (Tech Bulletin
`2005).
`
`“[I]t is important to optimize the amount of
`oligo that we need to immobilize on these
`double-attachment beads . . . in with
`address oligonucleotides (25mers),
`beads with assay sequences (50mers),
`and beads with both oligonucleotides
`immobilized together on the two attachment
`
`
`
`7
`
`
`
`Page 7 of 10
`
`

`
`
`
`Claim
`
`
`Claim Language
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`beads (see Figure 6, model C).” Steinberg
`2006 at p. 846.
`
`3.
`
`The bifunctional molecule of
`claim 1 wherein said polymer is
`an oligosaccharide, polypeptide,
`glycolipid, lipid, proteoglycan,
`glycopeptide or oligonucleotide.
`
`10.
`
`A library comprising a plurality of
`species of bifunctional molecules
`according to claim 1.
`
`16.
`
`The library of claim 10 wherein
`each of said species of
`bifunctional molecules in said
`plurality is present in molar
`equivalents of from 0.2 to 10.0.
`
`Model C of Fig. 6 at p. 845.
`The BeadArray product uses all of the
`limitations of Claim 1, see Claim 1[1] to 1[6].
`The BeadArray product also uses a polymer
`that is an oligonucleotide.
`
`See also Claim 1[2].
`The BeadArray product uses all of the
`limitations of Claim 1, see Claim 1[1] to 1[6].
`The BeadArray product also uses a library
`comprising a plurality of species of
`bifunctional molecules polymer [i.e., bead
`types] that is an oligonucleotide.
`
`“For very large numbers of bead types
`(tens to hundreds of thousands), a more
`practical decoding approach uses two
`different address oligos on each bead. This
`reduces the number of decoder oligos
`required to two times the square root of the
`number of unique beads. For example, to
`decode 1 million different bead types
`would require only 2000 oligo decoders.” Id.
`at p. 841 (emphases added).
`The BeadArray product uses all of the
`limitations of Claim 10, see Claim 10, which
`depends on Claim 1, see Claim 1[1] to 1[6].
`The BeadArray product also uses a plurality
`of species that are present in molar
`equivalents of from 0.2 to 10.0.
`
`“Illumina’s novel approach (6) enables the
`production of randomly assembled arrays in
`which the location of a probe is initially
`
`
`
`8
`
`
`
`Page 8 of 10
`
`

`
`
`
`Claim
`
`
`Claim Language
`
`Steinberg et al., BIOCONJUGATE CHEMISTRY
`17:841-848, 2006 (“Steinberg 2006”)
`unknown (7).” Id. at p. 841 (emphasis
`added). Steinberg 2006 cites reference 6
`(“Gunderson 2004”) to describe Illumina’s
`novel approach. With respect to molar
`equivalents, Gunderson 2004 describes
`Illumina’s BeadArray product having ~30-
`fold average redundancy with a minimum of
`5-fold redundancy, representing a molar
`equivalents range of 6.0.
`
`“The error rate of <1 X 10-4 per bead has a
`negligible impact on assay accuracy
`because of the ∼∼∼∼30-fold average
`redundancy and fivefold minimum
`redundancy of each bead type (Fig. 4).”
`Gunderson 2004 at p. 873 (emphasis
`added).
`
`
`
`Fig. 4 at p. 874.
`
`With regard to the process claims in U.S. Patent 5,573,905, we will withdraw this patent
`from the negotiations for the time being based on your representation that Illumina does
`not use PCR amplification in the decoding process, subject to further analysis and
`review of Illumina’s BeadArray manufacturing process, if necessary. As for any alleged
`conversations in Tacoma during the Syntrix v. Illumina trial, I do not know what you are
`referring to. Please be specific about what was allegedly said in these conversations, to
`whom, and when. I can represent that I was not engaged by Scripps during the Syntrix
`trial, that I had not done any analysis of the ‘905 and ‘596 patents at that time, and that I
`did not and could not have put you on notice concerning these patents. And, as far as
`notice is concerned, we’d like to know when and how Illumina first became aware of the
`Scripps patents.
`
`
`
`9
`
`
`
`Page 9 of 10
`
`

`
`
`
`In an effort to be cooperative and completely transparent, Scripps has taken the
`extraordinary step of providing you with a detailed analysis of the relevant claims and
`how Illumina’s Sentrix and BeadArray products meet the requirements of the specific
`claim language. As a result of our having provided this information, we hope you will
`recognize the many benefits of further discussing these issues and working towards a
`cooperative resolution. Again, we believe there should be a process more efficient than
`litigation that will enable us to resolve this dispute. Given that both parties are located
`within a few miles of each other, perhaps a face-to-face meeting would be beneficial.
`We look forward to hearing from you further.
`
`Sincerely,
`
`Jeff Oster, Esq. (mobile: 206 713 5467)
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`cc: Tom Northrup, Esq., Scripps
` Tom Fitting, Esq., Scripps
` Derek Gilliland, Esq., Nix Patterson & Roach, LLP
` Herbert J. Hammond, Esq., Thompson & Knight, LLP
`
`
`
`10
`
`
`
`Page 10 of 10