Patent: 5,728,677

Summary for Patent: 5,728,677

Title:	Methods of inhibiting T-cell dependent proliferation of peripheral blood lymphocytes using the CD2-binding domain of lymphocyte function associated antigen 3
Abstract:	Polypeptides and proteins comprising the CD2-binding domain of LFA-3 are disclosed. DNA sequences that code on expression for those polypeptides and proteins, methods of producing and using those polypeptides and proteins, and therapeutic and diagnostic compositions are also disclosed. Deletion mutants unable to bind CD2 and methods for their use are also disclosed. In addition, fusion proteins which comprise the CD2-binding domain of LFA-3 and a portion of a protein other than LFA-3, DNA sequences encoding those fusion proteins, methods for producing those fusion proteins, and uses of those fusion proteins are disclosed.
Inventor(s):	Wallner; Barbara P. (Cambridge, MA), Miller; Glenn T. (Haverhill, MA), Rosa; Margaret D. (Winchester, MA)
Assignee:	Biogen, Inc. (Cambridge, MA)
Application Number:	08/459,512
Patent Claims:	see list of patent claims

United States Patent 5,728,677: Claims, Critical Validity Review, and US Patent Landscape

US Patent 5,728,677 covers methods that inhibit T cell-dependent proliferation of peripheral blood lymphocytes by administering CD2-binding polypeptides (and related fusion proteins with a non-LFA-3 C-terminal domain, often Fc). The claims are tightly centered on (i) specific CD2-binding sequence constructs derived from an LFA-3-derived architecture and (ii) functional inhibition of T cell proliferation.

What do the core claims actually require?

Independent claim 1: polypeptide administration with CD2-binding capability

Claim 1 is a method claim with these required elements:

1. Administer to a mammal a polypeptide defined structurally as:
   • X1 - X2 - (SEQ ID NO:1) Asn Arg Val Tyr Leu Asp Thr Val Ser Gly - Y
2. X1 (if present) is hydrogen or methionyl.
3. X2 (if present) is a polypeptide consisting of the carboxy-terminal 1 to 77 amino acids of SEQ ID NO:5 (the claim then lists the full 1 to 77 region).
4. Y is hydroxyl or a polypeptide consisting of the amino-terminal 1 to 32 amino acids of SEQ ID NO:33 (sequence listed in full).
5. The polypeptide must be capable of binding to CD2.
6. The method outcome is inhibiting T cell-dependent proliferation of peripheral blood lymphocytes.

Interpretation pressure points

• The claim is drafted as a sequence-defined CD2-binding inhibitory construct. The functional language (“capable of binding to CD2”) is a potential evidentiary bridge, but the scope is already constrained by sequence requirements.
• The use of “X1 if present” and “X2 if present” makes the claim flexible at the ends while staying rigid in the central sequence logic.

Dependent claim 2: narrower X2 and shorter Y

Claim 2 narrows:

• X2 is limited to carboxy-terminal 1 to 50 amino acids of SEQ ID NO:2 (sequence listed).
• Y is limited to amino-terminal 1 to 10 amino acids of SEQ ID NO:3 (sequence listed).
• Still requires CD2 binding and T cell proliferation inhibition.

This dependent claim strengthens enforceability against close variants because it pins down shorter termini that could otherwise be argued as “equivalents.”

Dependent claim 3 and claim 4: explicitly listed sequence alternatives

Claim 3 selects polypeptides by sequence ranges or specific entries:

• amino acids 29-120 of SEQ ID NO:10
• amino acids 29-108 of SEQ ID NO:10
• amino acids 48-108 of SEQ ID NO:10
• SEQ ID NO:7

Claim 4 then defines SEQ ID NO:7 precisely:

• Lys Asn Arg Val Tyr Leu Asp Thr Val Ser Gly Ser Leu Thr Ile Tyr

These claims remove the need to prove full structural compliance for the listed alternatives: the patent holder can argue that infringement is met if the accused product matches one of the enumerated sequence blocks.

---

Independent claim 5: same targeting idea, but as a fusion protein

Claim 5 recasts the therapy as a fusion protein that includes:

1. An amino-terminal region consisting of one of the CD2-binding polypeptides listed in subparts (a) to (f), including:
   • the broad X1-X2-(SEQ ID NO:1)…-Y format with CD2-binding requirement
   • specific shorter X2/X2 and Y variants
   • amino acid windows in SEQ ID NO:10 (29-120, 29-108, 48-108)
   • SEQ ID NO:7
2. A carboxy terminal region that comprises a domain of a protein or polypeptide other than LFA-3.

Functional consequence

• Like claim 1, the fusion protein is administered to inhibit T cell-dependent proliferation.

Dependent claims 6 to 9: Fc-centric fusion protein architecture

Claim 6 adds:

• The non-LFA-3 carboxy-terminal domain comprises at least a portion of the Fc region of an immunoglobulin.

Claim 7 limits to:

• hinge region + CH2 and CH3 constant domains.

Claim 8 specifies hinge capable of forming intermolecular disulfide bonds plus CH2/CH3.

Claim 9 fixes an example length:

• fusion protein has amino acids 29-347 of SEQ ID NO:43.

Critical claim consequence

• The architecture reads like a CD2-binding “soluble decoy” fused to Ig Fc to gain:
  • dimerization via hinge disulfides (if the claim elements are used)
  • extended half-life via Fc engagement
  • multivalent avidity that can improve functional potency (although the claims do not expressly require mechanism, only inhibition and CD2 binding constructs).

---

What is the technical and legal “center of gravity” of the patent?

Claim scope is anchored in CD2-binding sequence constructs

Across claims 1 to 5, the essential invention is not broad “T cell immunosuppression.” It is sequence-defined CD2-binding polypeptides plus optional fusion to an Fc or other domain.

The claims therefore live at a specific intersection:

• CD2 as the receptor ligand target
• LFA-3-derived or LFA-3-like structural elements used to bind CD2
• fusion format to improve pharmacology

What would be hardest to design around?

Based on the claim drafting, the hardest non-infringing path is to build a competitor that still:

• uses a CD2-binding polypeptide with substantially the same binding epitope derived from the enumerated sequences/ranges, while also
• offering a comparable inhibitory effect

Claim 3 and claim 5 explicitly list multiple windows and SEQ ID NO entries, reducing room for “almost-the-same” design.

What could be easier to design around?

The clearest design-around levers, purely from claim language, are:

• avoid the listed specific sequence windows (claim 3 / claim 5 subparts)
• avoid the exact terminal constructions constrained by “amino-terminal 1 to 32” / “1 to 10” and “carboxy-terminal 1 to 77” / “1 to 50” definitions
• avoid the fusion format requiring “carboxy terminal region comprises a domain other than LFA-3” if a different overall architecture is used (though a fusion is not the only structure in the patent: claim 1 still covers non-fusion polypeptides)

---

Critical validity analysis: likely attack surfaces and claim weaknesses

1) Broad functional language tied to tight sequence

The patent pairs narrow sequence constraints with broad functional language (“capable of binding to CD2” and inhibition of proliferation). This is usually defendable if the specification supports binding and functional inhibition.

A validity challenge would target whether:

• “capable of binding to CD2” is supported across the entire stated sequence universe
• the claimed constructs are actually defined such that a POSA could make them and verify CD2 binding and inhibition without undue experimentation

But the claim itself is sequence-defined with enumerated windows; that tends to reduce “functional claiming without adequate structure.”

2) Potential indefiniteness arguments are limited

The claim defines sequences and termini numerically, and states binding capability. Indefiniteness typically arises where:

• functional limitations are the only boundary or
• sequence definitions are ambiguous in mapping to actual molecules

Here, the sequence definitions are explicit enough to avoid common indefiniteness traps.

3) Written description and enablement risks depend on breadth

If the specification only demonstrates activity for a subset of constructs, a challenger can argue that:

• not all sequence variants are enabled to inhibit proliferation
• CD2 binding is not shown for every construct across the claimed windows

This risk is most relevant to claim breadth created by:

• “X1 if present”
• “Y is hydroxyl OR a polypeptide of amino-terminal 1 to 32”
• flexible “X2 if present” constructs
• the “portion thereof” phrasing in multiple places

Still, because the claim limits are sequence windows rather than arbitrary portions, a written description challenge is not automatically compelling.

4) Obviousness exposure in light of CD2 biology and soluble decoy precedents

A common validity theme for immuno-oncology biologics patents is that:

• receptor-ligand decoy strategies were known
• Fc fusion to extend half-life and enable dimerization was known
• CD2 as an immunological checkpoint was known
• LFA-3 and CD2 interaction biology was known

For US 5,728,677, the most likely obviousness argument would combine:

• prior art LFA-3/CD2 interaction reagents or fragments
• prior art Fc fusion approaches for immune modulation
• rational engineering to create CD2-binding soluble constructs

Whether that combination succeeds depends on whether prior art already disclosed constructs matching:

• the specific SEQ ID windows and
• the same inhibitor strategy against T cell proliferation

If prior art used different epitopes, different truncation boundaries, or different fusion logic, the patentee can argue the claims are non-obvious.

5) Claim differentiation and scope stacking

Claims 1 and 5 are parallel: polypeptide vs fusion protein. Dependent claims 2, 3, 4 and 6 to 9 progressively narrow scope.

This can help enforceability if a court reads the structure limitations strictly. It can also create an obviousness narrative if the dependent narrowing is framed as predictable truncation and predictable Fc fusion.

---

US patent landscape: where the battle likely sits

A full claim chart and family-by-family citation audit cannot be completed from the prompt alone because US 5,728,677 depends on:

• its application history
• priority date and prosecution record
• the specification’s exact disclosure support for each sequence
• external prior art references that must be searched and tested against the claim elements

With only the claims text provided, a defensible landscape can be stated only at the level of landscape themes rather than a mapped list of patent numbers.

Landscape clusters relevant to this patent

1. CD2 axis immune modulators
   • antibodies, soluble receptors, ligand decoys, and fragments targeting CD2/CD58/LFA-3 interactions.
2. LFA-3-derived CD2-binding fragments
   • engineered polypeptide fragments that bind CD2 to modulate T cell activation and proliferation.
3. Fc-fusion immunomodulators
   • hinge-containing Fc fusions that dimerize and extend half-life.
4. T cell proliferation inhibition therapies
   • broad immunosuppressive method claims that become narrow when tethered to specific CD2-binding sequences.

Where 5,728,677 likely sits in that landscape

• The patent is not a general “CD2 pathway inhibition” patent. It is a sequence-defined CD2-binding fragment/fusion strategy aimed specifically at inhibiting T cell-dependent proliferation of peripheral blood lymphocytes.
• That places it in a “defensible specificity” zone: the key competitive threat is molecules that also use LFA-3-like CD2 binding sequences and similar truncation boundaries, especially when fused to Fc.

---

Practical claim-to-product risk assessment

What products are most at risk

Any US-launched competitor that has:

• a CD2-binding polypeptide that falls within the enumerated SEQ ID windows/ranges (claim 3) or matches the X1/X2/SEQ ID NO:1/Y architecture with the exact defined truncations (claims 1 and 2), and
• includes inhibition of T cell-dependent proliferation in its intended use or labeling.

What competes with the patent but can escape it

• CD2-binding agents that use different binding determinants not contained in the claim-defined windows.
• CD2 modulation agents that are antibodies or formats that do not match the required “administer a polypeptide/fusion protein having the defined sequences” language.
• approaches that inhibit T cell proliferation through CD2-independent pathways while avoiding the defined constructs.

---

Key Takeaways

• US 5,728,677 is anchored on sequence-defined polypeptides that bind CD2 and inhibit T cell-dependent proliferation of peripheral blood lymphocytes.
• The most enforceable scope is created by enumerated sequence windows (claim 3) and the terminal truncation architecture in claims 1 and 2.
• Claims 5 to 9 add Fc-centric fusion structure constraints (hinge capable of disulfide-mediated intermolecular bonding plus CH2/CH3).
• The core design-around lever is not the biological target but the specific sequence boundaries and fusion architecture required by the claims.
• The most likely validity challenges would be obviousness (CD2 decoys plus Fc fusion as known engineering) and enablement/written description across the full range of claimed constructs.

---

FAQs

1) Does the patent require a specific mechanism beyond CD2 binding?

No. The claims require “capable of binding to CD2” and functional inhibition of T cell-dependent proliferation. They do not require a single downstream mechanism.

2) Is an Fc fusion mandatory to infringe?

No. Claim 1 covers administering a CD2-binding polypeptide without requiring fusion. Claims 5 to 9 cover fusion proteins with non-LFA-3 C-terminal domains, including Fc variants.

3) How do claim 3 and claim 4 change enforcement strength?

They create explicit sequence alternatives. If an accused construct matches one of those enumerated windows, the sequence-mapping step becomes straightforward.

4) What are the two biggest technical boundary conditions for infringement?

(1) Matching the defined amino acid sequence/terminal truncation structures (X1, X2, Y windows and SEQ ID-based segments), and (2) demonstrating that the construct is a CD2-binding polypeptide used to inhibit the specified proliferation endpoint.

5) What is the most plausible competitor strategy to reduce infringement risk?

Use CD2 modulation without using the claim-defined sequence windows and truncations, or use an architecture that does not match the polypeptide/fusion protein sequence constraints tied to CD2 binding.

---

References (APA)

[1] US Patent 5,728,677. (n.d.). Claims text as provided in the prompt.

Details for Patent 5,728,677

Applicant	Tradename	Biologic Ingredient	Dosage Form	BLA	Approval Date	Patent No.	Expiredate
Astellas Pharma Us, Inc.	AMEVIVE	alefacept	For Injection	125036	January 30, 2003		2015-06-02

International Patent Family for US Patent 5,728,677

Country	Patent Number	Estimated Expiration
Austria	E193724
Australia	1754092
Australia	660981