Patent: 8,044,174

Summary for Patent: 8,044,174

Title:	Thrombopoietic compounds
Abstract:	The invention relates to the field of compounds, especially peptides or polypeptides, that have thrombopoietic activity. The peptides and polypeptides of the invention may be used to increase platelets or platelet precursors (e.g., megakaryocytes) in a mammal.
Inventor(s):	Liu; Chuan-Fa (Longmont, CO), Feige; Ulrich (Newbury Park, CA), Cheetham; Janet C. (Montecito, CA)
Assignee:	Amgen Inc. (Thousand Oaks, CA)
Application Number:	10/933,133
Patent Claims:	see list of patent claims

Critical patent-claims analysis and competitive landscape for US Patent 8,044,174 (TMP1-Ln-TMP2 MPL receptor binder)

US 8,044,174 claims a class of MPL (myeloproliferative leukemia protein) receptor-binding peptidic constructs built from two “TMP” peptide blocks (TMP1 and TMP2) fused at the TMP1 C-terminus to the TMP2 N-terminus, optionally through a linker L1, with extensive amino-acid/sequence selection options and allowed spacer/linker chemistries. The claim set is broad on sequence diversity (multiple families for TMP1 and TMP2, optional linker presence/length, and linker composition subsets) and broad on modification chemistry (non-peptidyl bond replacements and derivatization including N-/C-cap functionalization), but it is still constrained by the core architectural requirements: an MPL-binding construct of the TMP1-(L1)n-TMP2 format and TMP1/TMP2 drawn from the specific enumerated “core compound” sub-structures.

Because the claim text you provided is only claim 1–8 (and does not include the patent’s specification definitions for “TMP,” the exact mpl-binding assay thresholds, or the full list of dependent claims), the landscape assessment below focuses strictly on the claim scope, likely freedom-to-operate (FTO) pressure points, and the litigation/validity-risk vectors that flow from the claim mechanics and breadth.

---

What does US 8,044,174 claim protect: TMP1–(L1)n–TMP2 MPL receptor binders?

Direct answer: The patent protects MPL-binding compounds whose primary structure matches TMP1-(L1)n-TMP2, where TMP1 and TMP2 are each selected from enumerated 9- to 10-residue “core” amino-acid patterns (X2–X10) that must satisfy positional residue sets, with optional linker chemistry and optional stereochemical variants (D-amino acids).

Claim 1 architecture: the fixed backbone + constrained combinatorics

Claim 1 is structured as follows:

1. Binding constraint: “a compound that binds to an mpl receptor” (functional limitation).
2. Structural constraint: TMP1-(L1)n-TMP2, where TMP1 C-terminus is linked to TMP2 N-terminus, optionally via L1.
3. Linker constraint:
   • n is 0 or 1
   • if n=1, L1 is “independently selected from linker groups consisting of Y^n,” with Y a naturally occurring amino acid or stereoisomer, and n is 1 through 20.
4. Peptide constraints: TMP1 and TMP2 are each independently selected from one of eight “core compound” families (a–h) with explicit residue-position selection rules for X2–X10.
5. Stereochemistry and salts: “physiologically acceptable salts thereof.”

The key scope-driving elements

• TMP1/TMP2 linkage is mandatory. Changing the fusion site (for example, using internal linking or altering terminal amide/peptide bond type without falling under the “derivatized” alternatives) is an obvious design-around lever.
• Sequence diversity is large. The claim does not specify fixed residues for most positions; it enumerates amino-acid sets at each position. That tends to expand coverage to many analogs.
• Linker presence is optional but bounded. n is only 0 or 1; if present, L1 length can be 1–20 residues, which is another breadth amplifier.
• Functional “binds to mpl receptor” converts some uncertainty into litigation fuel. Infringement can turn on whether an accused compound meets binding, even if sequence falls within the scaffold.

Claimed peptide “core families” (a–h): what matters for infringement

All eight families share a repeating motif: X4 = Pro, X5 is Thr/Ser (or fixed Ser in some families), X6 is hydrophobic/aromatic-rich sets, X7 is basic (Arg/Lys), X8 includes Gln/Asn/Glu, and X9 and X10 are aromatic/sulfur and hydrophobic sets. The differences among a–h largely shift polar/basic composition at X2–X3–X6–X9–X10 and whether X7 is Lys only (in family e) or Arg/Lys (in others), and whether X9 is restricted (e.g., family g has X9 limited to Tyr/Cys/Ala/Phe).

That pattern is consistent with a claim strategy: lock in a pharmacophore-like distribution while leaving room for analog variation.

---

How broad are the sequence options in claim 1: how many TMP1/TMP2 variants exist?

Direct answer: The claim enumerates positional amino-acid sets for X2–X10 across eight families, but without converting to numeric cardinalities for each position (because some families list multiple options and some positions are singleton), a reliable exact count of distinct peptides cannot be computed from the claim snippet alone. What can be stated precisely is that the combinatorics are multiplicative across positions, and TMP1 and TMP2 are selected independently, making the number of possible TMP1/TMP2 combinations extremely large.

Breadth multipliers

• Independent selection for TMP1 and TMP2: even if TMP1 has N possible variants and TMP2 has M, total scaffolds scale roughly as N×M.
• n=0 vs n=1: the presence/absence of a linker introduces an additional axis.
• Linker length 1–20: if L1 is a peptide linker built from amino acids, the sequence space for linker combinations grows superlinearly.
• Physiologically acceptable salts: adds additional infringement targets if the core compound is present in salt form.

---

What do the dependent claims add: extended TMP lengths, derivatizations, and D-amino-acid coverage?

Claim 2: expanded TMP1/TMP2 patterns with X11–X14 tails (bigger peptidic scope)

Claim 2 broadens the structure of TMP1 and TMP2 to include additional residues:

• TMP1/TMP2 are independently selected from patterns containing X1–X14 combinations where X2–X10 are as defined in claim 1, and X1 and X11–X14 are selected from specified amino-acid sets.
• The enumerated patterns include:
  • X2–X10–X11 (11 residues segment for one end)
  • X2–X10–X11–X12 (12)
  • X2–X10–X11–X12–X13 (13)
  • X1–X2–X10 (10 plus N-terminal variable)
  • up to X1–X2–X10–X11–X12–X13–X14 (14 total positions)

Claim 2 therefore extends protection to longer constructs and N-/C-terminal residue variability consistent with claim 1’s core pharmacophore.

Claim 3: non-peptidyl linkage replacements and N-/C-cap functionalization (chemistry design-around risk)

Claim 3 adds a major infringement lever: it explicitly covers replacements of peptidyl [–C(O)NR–] bonds with non-peptidyl linkages, including:

• –CH2–carbamate (–CH2–OC(O)NR–)
• phosphonate
• –CH2–sulfonamide (–CH2–S(O)2NR–)
• urea (–NHC(O)NH–)
• –CH2–secondary amine
• alkylated peptidyl linkage (–C(O)NR6 where R6 is lower alkyl)

It also covers N-terminus and C-terminus modifications, such as:

• N-terminus: –NRR1, –NRC(O)R, –NRC(O)OR, –NRS(O)2R, –NHC(O)NHR (with a proviso)
• N-protection/capping equivalents including CBZ-based groups and substituted phenyl rings
• C-terminus: –C(O)R2 where R2 is lower alkoxy or –NR3R4

For a freedom-to-operate analysis, this claim language compresses a typical design-around strategy (changing amide bonds and terminal caps) into a smaller-safe zone because the patent pre-authorizes many alternative bond types as still within claim coverage.

Claims 4 and 5: full or partial D-configuration

• Claim 4: “all amino acids have a D configuration.”
• Claim 5: “at least one amino acid has a D configuration.”

This matters because many analogs are made as D-peptide variants for stability and protease resistance. The patent anticipates that both full and partial epimerization still fall within scope.

Claims 6–8: linker L1 specificity with examples, including Cys and defined sequences

Claim 6 constrains L1 as ((Gly)n, n=1–20) with substitution rules (up to half of glycines replaced with another natural amino acid or stereoisomer). Claim 7 provides explicit L1 examples (SEQ IDs 6–9). Claim 8 adds L1 comprises a Cys residue.

Net effect: the patent claims both generic glycine-rich linkers and specific linker exemplars including cysteine.

---

How strong is US 8,044,174 as a patent estate: what claim traits drive enforceability vs invalidity?

Direct answer: Enforceability strength is driven by (i) clear structural scaffold definition (TMP1-(L1)n-TMP2), (ii) extensive but bounded enumerations of amino-acid sets, and (iii) explicit chemical-derivatization coverage in claim 3. Validity risk typically increases where functional “binds to mpl receptor” is broad relative to disclosure, and where the “linker groups consisting of Y^n” language may be interpreted broadly without clear structural anchors. The net effect is a patent that can be powerful in litigation because it targets both primary sequences and common chemistry modifications, but faces attack if the disclosure does not support binding across the full enumerated scope.

Where infringement usually lands quickly

• If an accused compound matches TMP1 and TMP2 residue-position restrictions and maintains the terminal fusion geometry, infringement is structurally plausible.
• If the accused molecule uses non-peptidyl linkages or alternative caps that map onto claim 3 categories, infringement risk remains high.
• D-peptide analogs are not a safe harbor due to claims 4 and 5.

Where infringement can become contestable

• The functional clause: “binds to an mpl receptor.”
• If binding is not demonstrated for a particular analog (or if assay conditions differ), parties may dispute whether the compound meets the functional limitation.
• If an accused compound has sequence substitutions outside the enumerated sets for X2–X10 or X1/X11–X14, it may avoid the claim’s structural predicate even if it binds MPL.

---

What are the likely design-around options versus US 8,044,174?

Direct answer: Design-around tends to work when it breaks one of the claim’s three hard constraints: the TMP1/TMP2 core families at specified residue positions, the TMP1 C-terminus to TMP2 N-terminus fusion requirement, or the covered linker/derivatization categories.

High-leverage design-around moves

1. Break the scaffold: use a different fusion topology that does not fit TMP1-(L1)n-TMP2 with TMP1 C-terminus linked to TMP2 N-terminus.
2. Escape positional residue sets: introduce residues at X2–X10 (or X1/X11–X14 if claim 2 matters) that are outside the enumerated options for all families a–h and the dependent family constraints.
3. Use linkers not falling under claim 1/6 definitions: claim 1 allows L1 as amino-acid-based Y^n with n 1–20, and claim 6 tightens to glycine-rich linkers substituted up to half. Linkers that are clearly non–amino-acid-based (or structurally not captured by “linker groups consisting of Y^n”) may reduce risk, though claim 3 covers several bond-type replacements.
4. Avoid D-configuration coverage if that is used as an assumed escape route: D-configuration is explicitly covered, so stereochemical switching is not protective by itself.

---

Which competitors are at risk from this claim scope: what product classes would plausibly overlap?

Direct answer: Any MPL receptor agonist peptidic construct that matches the TMP1/TMP2 architecture, uses similar core residue distributions, and preserves terminal fusion geometry is the main at-risk zone. This includes analogs aiming for protease stability (D-amino-acids) and analogs that adjust linker chemistry (carbamate/urea/sulfonamide/phosphonate) because claim 3 anticipates those.

Where overlap is most likely

• Development programs that copy the same scaffolding but alter:
  • linker length/composition (gly-rich and Cys-containing linkers)
  • bond types (amide to carbamate/urea/sulfonamide/phosphonate)
  • terminal caps
  • D/L epimerization patterns

---

What does this mean for “generic” or “biosimilar” entry risk?

Direct answer: The patent reads like a small-molecule/peptide-like composition-of-matter for an MPL agonist binder, not a biologic in the conventional mAb sense. There is no “generic” pathway for such peptides in the way there is for small-molecule drugs. The entry risk instead manifests as copycat analogs that attempt to claim/market a functionally similar MPL agonist with different sequences or chemistry.

If a biosimilar analogy is used, the same issue applies: peptide MPL agonists typically require bridging for structural differences, and patent risk is driven by whether the entry compound stays within the structural families and derivatization categories.

---

What litigation issues typically arise from this claim set: claim construction pressure points

Direct answer: Expect disputes around (i) whether “mpl receptor” binding is sufficiently defined, (ii) whether “TMP” and “core compound” definitions are limited to specific sequences in the specification or read broadly to cover the entire enumerated residue-set space, and (iii) whether accused linkers/bonds fall within claim 3’s non-peptidyl linkage replacements and terminal modifications.

Construction pressure points

• “TMP.sub.1” and “TMP.sub.2” definitions: if the specification ties TMP1/TMP2 to particular example sequences or structural definitions, that can narrow or broaden the interpretation.
• “linker groups consisting of Y^n”: breadth depends on how “Y^n” is construed. If interpreted as “peptide-like amino-acid linker,” scope is broad. If the specification restricts Y to a specific set, scope narrows.
• D-amino acid coverage: if a compound has a mix, claim 5 captures at least one D residue, making stereochemistry a non-exit.

---

Key takeaways

• US 8,044,174 protects MPL-binding peptidic constructs built on TMP1-(L1)n-TMP2 with terminal fusion geometry and enumerated positional amino-acid sets for core segments.
• The dependent claims expand scope to longer TMP termini (X1/X11–X14), cover major bond-type derivatizations (carbamate/urea/sulfonamide/phosphonate/alkylated amide equivalents and multiple terminal caps), and include D-amino-acid variants.
• The strongest infringement triggers are compounds that preserve the scaffold while modifying linker length/composition and bond chemistry for stability.
• Design-around must typically change scaffold topology or escape the enumerated residue-position sets and/or covered linker/derivatization categories.

---

FAQs

1) What is the core compound in US 8,044,174: which positions are fixed and which vary?

The core is the X2–X10 pattern across families a–h, with positional constraints such as X4 = Pro and X7 = basic (Arg/Lys or Lys-only in family e), while other positions vary within specified amino-acid sets.

2) Does US 8,044,174 cover D-peptides made for stability?

Yes. The patent explicitly covers constructs where all amino acids are D (claim 4) and where at least one is D (claim 5).

3) Can a competitor avoid infringement by converting an amide bond to a carbamate or urea?

Not by default. Claim 3 lists several non-peptidyl replacements, including carbamate, phosphonate, sulfonamide, and urea, as within the claim’s coverage.

4) Is the linker required in US 8,044,174?

No. Claim 1 sets n=0 or 1, so the linker can be absent or present. If present, its length is capped by n=1–20.

5) What turns the functional “binds to mpl receptor” limitation into litigation risk?

Even if a compound matches the structural scaffold, parties may dispute whether it demonstrably binds MPL under relevant assay conditions, especially for analogs spanning the full enumerated sequence and linker space.

---

References

No external sources were cited because the analysis is derived solely from the claim text provided in the prompt.

Details for Patent 8,044,174

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Amgen Inc.	NPLATE	romiplostim	For Injection	125268	August 22, 2008	8,044,174	2024-09-02
Amgen Inc.	NPLATE	romiplostim	For Injection	125268	July 22, 2019	8,044,174	2024-09-02

International Patent Family for US Patent 8,044,174

Country	Patent Number	Estimated Expiration
Argentina	020934
Austria	E348163
Australia	1223900
Australia	773891
Bulgaria	105401