US 6653098

I'll search for authoritative information on this patent number and any related litigation.

Let me run a couple more targeted searches to check specifically for litigation or CAFC docket activity involving this patent.

US Patent 6,653,098 — Analyst Summary

Scope note: I retrieved the full authoritative patent text (Google Patents, fetched 2026-09-24) and searched for USPTO/CAFC activity. I found no CAFC 2026 docket entry and no litigation record involving patent 6,653,098 — see the "Litigation / CAFC check" section below for details and caveats.

Bibliographic data

Small caveats on the record itself:

• The assignment record lists assignors as Casperson, Gerald F.; Polazzi, Joseph O.; and Violand, Bernard N. — which is not identical to the two named inventors (Violand, Harding). The WO pamphlet counterpart lists inventors as Harding, E.I. and Violand, B.N. I'm reporting both as they appear rather than reconciling them.
• In the sequence-listing tables, SEQ ID NOs 10 and 11 are both tagged "PRT / human" in the machine-generated metadata, yet the specification text defines SEQ ID 10 as mouse endostatin and SEQ ID NO 11 as human endostatin. This looks like a sequence-listing annotation artifact; flagged, not corrected.

Abstract (verbatim)

"Methods for producing mouse and human endostatin are disclosed. Methods for refolding and purifying endostatin from inclusion bodies expressed in bacteria and nucleic acids encoding full-length and truncated forms of endostatin are also disclosed."

Technical gist

The patent addresses a real problem in the endostatin field: endostatin expressed in E. coli forms insoluble inclusion bodies, and prior work (WO 97/15666 / O'Reilly et al.) used the material as an unrefolded suspension because attempts to refold it lost >99% of protein. This patent claims a specific solubilization + refolding + isolation regime that yields soluble, properly folded, biologically active material at useful yields.

Core process parameters described:

• Solubilize inclusion bodies at elevated pH (~9–11.5; preferably ~10.5) with a denaturant (urea ~4–10 M, pref. 6 M; or GuHCl ~2–8 M) plus a reducing agent.
• Refold at near-neutral pH (~6–8.5; pref. ~7.5) at lower denaturant (urea ~2–4 M, pref. 3.5 M; or GuHCl ~0.2–2 M), in the presence of a redox interchange agent (cystine/oxidized glutathione), with disulfides formed via air oxidation over ~12–96 h (pref. ~60 h).
• Purify via acid precipitation + ion exchange (S-Sepharose), or other chromatography (HIC, RP-HPLC).

Working examples: mouse endostatin dissolved in 6 M urea / 5 mM DTT / 50 mM Bis-Tris propane pH 10.8 at 2.5 mg/mL, then diluted to 0.25 mg/mL in 3.5 M urea, pH 7.0 to refold; human endostatin dissolved in 6 M urea / 10 mM cysteine, pH 10.8, then diluted to 0.25 mg/mL in 3.0 M urea, pH 7.5. Activity is shown via endothelial cell proliferation and migration assays (mouse: HMEC/CPAE; human: proliferation at 10 µg/mL).

Independent claims — plain-language overview

There is exactly one independent claim (claim 1); claims 2–26 are all dependent on claim 1.

Claim 1 — the sole independent claim

A method of producing endostatin, comprising:

1. (a) transforming a prokaryotic host cell with an expression vector carrying an isolated polynucleotide encoding endostatin;
2. (b) culturing the cells so that inclusion bodies containing endostatin form;
3. (c) recovering those inclusion bodies;
4. (d) solubilizing the endostatin at elevated pH in the presence of a denaturant;
5. (e) properly refolding the endostatin at about neutral pH; and
6. (f) isolating properly folded, active endostatin.

Numeric/chemical limitations built into claim 1:

• Elevated (solubilization) pH: pH 9 to about 11.5.
• Near-neutral (refolding) pH: pH 6 to about 8.5.
• Denaturant is urea: ~4–10 M during solubilization and ~2–4 M during refolding.
• Refolding is done in the presence of a disulfide-reducing agent, limited to: DTT at ~0.5–2 mM or cysteine at ~0.5–4 mM.
• Endostatin concentration during refolding: ~0.2–2 mg/mL.

Notable drafting point: claim 1 tells you the reducing agent must be present during refolding but leaves the solubilization-step reductant open (the specification's examples use 5 mM DTT or 10 mM cysteine there). Also, claim 1 limits the denaturant to urea, excluding the guanidine-HCl alternative that the specification describes as an embodiment.

Dependent claim groupings (all depend from claim 1)

• pH refinements: 2–5 (solubilization pH 10–11, ~10.5; refold pH 7.0–8.0, ~7.5).
• Urea concentrations: 6, 7 (6 M solubilization; 3.5 M refold).
• Reducing agent concentrations: 8–13 (DTT 2–10 mM / ~5 mM solubilization, ~0.5 mM refold; cysteine 5–20 mM / ~10 mM solubilization, ~1 mM refold).
• Air oxidation / disulfide formation: 14–17 (12–96 h; 24–72 h; ~60 h).
• Protein concentration: 18–20 (~1–20 mg/mL, pref. 2.5 mg/mL, at solubilization; ~0.25 mg/mL at refold).
• Downstream purification: 21 (ion-exchange, HIC, or RP-HPLC).
• Sequence/identity limitations: 22–23 (mouse endostatin; polynucleotide of SEQ ID NOs 5, 6, 7 or 8); 24–25 (human endostatin; SEQ ID NO 9).
• Host: 26 (E. coli).

Note on "prior art / cited references"

The record lists a small cited-art set (6 references), notably WO 1997/015666 (Children's Medical Center Corp. — the foundational endostatin application), WO 1996/040736 (Celtrix — IGF-I/IGFBP-3 folding), EP 0312358 and EP 0393725 (disulfide-bond formation in recombinant proteins), and US 5,854,205. Several third-party later filings cite US 6,653,098 (e.g., an Amgen refolding family, Bolder Biotechnology refolding applications). The inventor list here includes three names, which is one discrepancy flagged above.

Litigation / CAFC 2026 check — result and uncertainty

• I searched for a CAFC 2026 docket or any court proceeding naming patent 6,653,098. No match was found. The CAFC argument calendar results returned unrelated matters (e.g., In re Motorola Solutions, No. 25-134; other 2025–2026 appeals), and the district-court results concerned other patents entirely.
• I also searched for infringement suits involving the number 6,653,098 / "6,653,098" tied to G.D. Searle / endostatin and found no litigation record in the results returned.
• Uncertainty: Public web search is not a substitute for a direct PACER/CAFC docket query or the USPTO Patent Center "Litigation" tab, which I cannot definitively query here. So I can state confidently that I found no evidence of a CAFC 2026 appeal involving this patent, but I cannot rule out that one exists or once existed. The patent expired for fee reasons in 2019, which makes 2026 appellate activity improbable.

Bottom line: US 6,653,098 is a 2003 G.D. Searle patent (Violand/Harding) claiming a specific high-yield method for solubilizing, refolding, and isolating E. coli-expressed mouse and human endostatin from inclusion bodies — one independent claim (claim 1) with 25 dependent claims — and I found no CAFC 2026 docket activity associated with it.

I'll research litigation involving US patent 6653098. Let me search multiple sources.

Let me search more targeted litigation databases.

Let me check litigation-specific databases directly.

Litigation Search: US Patent 6,653,098

Bottom line

I found no litigation involving US Patent 6,653,098. Based on the sources I searched, this patent does not appear to have been asserted in any U.S. district court, Federal Circuit appeal, PTAB proceeding, or ITC action.

What the searches covered

I searched litigation-oriented queries (Unified Patents portal pattern, "6,653,098" / "6653098" + lawsuit/patent infringement, Docket Alarm, Justia dockets, and general CAFC/PACER-style litigation terms). Results returned:

• No case captions, docket numbers, or judgments referencing this patent.
• The Google Patents record for US6653098B1 itself contains no "Litigation" section (which Google populates when it indexes court records) — only "Cited By," "Citations," and family members. This is consistent with the patent never being asserted.
• The "Cited By" and "Families Citing" entries (e.g., Bolder Biotechnology, Biogen Idec, Amgen filings) are patent-document citations, not litigation.

The patent's identity (for the record, per your no-auto-correct rule)

Searches also surfaced unrelated items I am not conflating with this patent:

• JP Patent No. 6653098 — a Japanese patent on a sulfur-containing polymer (different patent, coincidentally same digits).
• RU 2278688 C1 and similar foreign documents — these merely cite US 6,653,098 as prior art; citation is not litigation.

Caveats

• My litigation searches returned mostly noise (unrelated NPE/Unified Patents matters), and three follow-up docket-specific queries (Justia dockets, Docket Alarm, ANDA/BPCIA phrasing) could not be completed within the tool-step limit, so coverage there is incomplete.
• Accordingly, my conclusion is: no litigation known or found — not a categorical assertion that none exists.

If a definitive negative is required, the authoritative checks are: PACER Case Locator / CourtListener RECAP (search "6653098"), the Unified Patents litigation portal (portal.unifiedpatents.com/patents/patent/6653098), and USPTO PTAB records. I'd recommend running those directly rather than relying on this summary.

I'll verify against live sources before finalizing, since the structured block indicates no proceedings on file.

PTAB / AIA Trial Activity — US Patent 6,653,098

Proceedings overview

Total AIA trial proceedings on file: 0 (zero). There is nothing to break down by status — no instituted IPRs, no PGRs, no CBMs, no denials, no terminations, no settlements, and no claim has ever been canceled or sustained in an AIA trial. The bottom line for a defendant: the patent has never been stress-tested at the PTAB, but that is now largely academic — the patent expired for fee reasons on 2019-01-14, so the relevant defensive posture is not "hardened by surviving IPRs" but rather "no live PTAB chessboard and, as of 2026, no realistic damages exposure." No IPR estoppel has ever attached to any party, so every invalidity ground remains available in district court — if anyone ever bothered to file there.

Source of record: the structured "PTAB proceedings on file" block supplied with this prompt, which states that the USPTO Open Data Portal API returns no AIA trial proceedings for this patent as of the most recent ingest. I independently searched for PTAB petitions against 6,653,098 and found none (see caveats).

Proceedings by proceeding

None to list. There is no proceeding number to report, and per the operating rules I will not invent one. For completeness, the fields requested would resolve as follows:

(no proceedings on file)

• Type: N/A
• Filed: N/A
• Status: N/A — absence of any record
• Judge panel: N/A
• Petition grounds: N/A
• Institution decision: N/A
• Final Written Decision: N/A
• Settlement / termination: N/A
• Appeal: N/A
• Defensive value: A defendant cannot rely on an earlier petitioner's work-product or an FWD's claim constructions, because neither exists. Conversely, a defendant also faces zero § 315(e)(2) estoppel — every ground is fresh. The practical defensive value is instead the expiration date: the patent's 20-year term ran out 2019-01-14, and any § 286 six-year damages lookback for conduct in the patent's life is now closed.

Strategic summary

Which claims are canceled vs. sustained vs. untested. Claim 1 — the sole independent claim, and therefore the load-bearing claim for all 25 dependents (claims 2–26) — remains untested at the PTAB, as do all dependents. Nothing was canceled and nothing was sustained; the claims have simply never been adjudicated in an AIA trial. That is an unusual posture for a patent that was commercially relevant in the anti-angiogenesis space and that sits atop a citation chain including Amgen and Bolder Biotechnology refolding families, but it is consistent with the fact that the patent's owner (G.D. Searle / GD Searle LLC) never asserted it — the earlier sections of this analysis found no district-court or Federal Circuit litigation either. A patent that is never asserted never attracts IPRs.

Estoppel landscape. Because no IPR or PGR was ever instituted, § 315(e)(2) estoppel is a null set. For any defendant facing assertion (hypothetically), that means no prior-art ground is foreclosed — patents, printed publications, and even § 102(a)(1)/(a)(2) public-use or on-sale art remain fully available, and the § 103 combinations that a panel never touched are all on the table. The natural invalidity targets remain the six references the examiner cited, most notably WO 1997/015666 (Children's Medical Center Corp., the foundational endostatin filing, which itself discloses E. coli expression and use of the unrefolded inclusion-body suspension) and EP 0312358 / EP 0393725 (disulfide-bond formation / activation of recombinant proteins in denaturant). Anyone revisiting § 103 would argue the claimed high-pH-solubilization / near-neutral-refolding window over that art. Note, though, that claim 1's combined numeric limitations (urea 4–10 M then 2–4 M; pH 9–11.5 then 6–8.5; reductant limited to DTT at 0.5–2 mM or cysteine at 0.5–4 mM; refold concentration 0.2–2 mg/mL) form a fairly tight cage, which is likely why the patent issued over the WO '15666 art in the first place.

Pattern signals. No repeat petitioner, obviously. No PTAB appeals by the patent owner. No defensive aggregator (Unified Patents, RPX, etc.) appears anywhere in the chain — the Google Patents "Cited By" and "Families Citing" entries (Bolder Biotechnology refolding applications, Biogen Idec/Amgen refolding families, a Chinese endostatin renaturation family) are patent-document citations, which are prior-art acknowledgments, not challenges. Do not mistake those citation lists for litigation or PTAB history.

Recommended next steps

1. If you are a defendant: there is no FWD to link to and no canceled claim to quote. Your invalidity theory is unconstrained by any prior PTAB record. Run your own § 102/§ 103 analysis against WO 1997/015666 and the EP disulfide-bond references, and note that the patent is expired — check whether any asserted theory covers conduct within the (§ 286) lookback window before it closed in 2025.
2. If you are the patent owner: there is no PTAB scare to manage. The constraints are commercial and temporal (expiration 2019-01-14), not procedural.
3. Verify, don't trust the summary. For a definitive negative, query:
   • USPTO PTAB E2E / Patent Trial and Appeal Board End-to-End and USPTO Patent Center → "Litigation" / "PTAB" tabs for Patent 6,653,098;
   • CourtListener RECAP and PACER Case Locator for "6653098";
   • Unified Patents litigation portal at portal.unifiedpatents.com/patents/patent/6653098.
     I could not run those portal queries directly here, so my result is "no proceedings found," not a categorical proof of none.

Caveats and flagged non-issues

• Do not confuse this patent with JP Patent No. 6653098 ("含硫黄ポリマー" / sulfur-containing polymer with ionic groups, application filed 2015-08-28, published 2016-04-14). Web searches for the bare string "6653098" surface this Japanese patent prominently. It is a different patent, different jurisdiction, different subject matter — it is search noise, not related to US 6,653,098.
• RU 2278688 C1 (Lutsenko et al., "Human endostatin preparation and method for production thereof") merely cites US 6,653,098 as an analogue/prior art. A citation is not a proceeding.
• Coverage limitation: my searches returned no PTAB docket entries, and the structured ODP block independently reports zero. Both point the same way, but neither is a substitute for a direct E2E/Patent Center query, so the correct formulation is "no PTAB activity found or on file", not a positive assertion that none has ever existed.

Bottom line: US 6,653,098 has zero AIA trial history — no IPR, PGR, or CBM was ever filed, instituted, or decided. Combined with its 2019-01-14 fee-related expiration and the absence of any litigation found in the earlier sections of this analysis, the patent is, in practical enforcement terms, dead weight: no claim is canceled, but no claim is enforceable going forward either.

I'll research the assignment record for US 6,653,098. Let me start with targeted searches.

Let me search for the specific assignment records and the company history.

Assignment & Ownership Reconstruction — US 6,653,098

Important sourcing caveat up front: I was able to retrieve the Google Patents legal-events record for US 6,653,098 and the corporate-history context, but I could not complete a live pull of the USPTO Assignment Center entry (assignmentcenter.uspto.gov / assignment.uspto.gov) before hitting the tool limit. Consequently I do not have the reel/frame, the correspondent-of-record, or the execution date for the one assignment that is visible. I am flagging this rather than inventing reel/frame numbers. Every reel/frame field below is marked [not retrieved] — that is a data gap, not a finding of "none."

Inventors

Unusual-pattern notes:

• Inventor/assignor mismatch (flagged, not corrected). The front page names two inventors (Violand, Harding). The recorded assignment's assignors are Casperson, Gerald F.; Polazzi, Joseph O.; and Violand, Bernard N. — i.e., Casperson and Polazzi assigned the case but are not named inventors, and named inventor Harding did not sign this particular recording. This is a real inconsistency in the record, not a transcription error on my part. Plausible benign explanations (a blanket/confirmatory assignment swept in co-workers; Harding's rights assigned under a separate employment instrument) cannot be confirmed from the sources reached, so I am reporting it as recorded.
• No evidence of a pre-filing inventor exodus. I found no basis to conclude the inventors departed Searle within 12 months of the Feb. 1998 priority / Jan. 1999 filing. The better-documented "departure" event is organizational, not individual: Pfizer's 2003 closure of the former G.D. Searle Skokie, IL site with ~1,500 layoffs/transfers, which post-dates issuance. That is corporate consolidation, not a portfolio fire-sale trigger.
• Note also Joseph O. Polazzi corroborates the Searle/Monsanto site: Polazzi is a co-author of an E. coli expression paper cited in this very specification (Allen, Polazzi, Gierse & Easton, J. Bacteriol. 174:6938-6947, 1992), i.e., the assignor set is the internal Searle expression-team, consistent with an employer-takes-all assignment rather than an arm's-length deal.

Original assignee

G.D. Searle & Co. (n/k/a G.D. Searle LLC) — the entity named on the issued patent.

• Line of business: Research-based pharmaceutical company (Chicago/Skokie, IL origins; later St. Louis-area R&D under Monsanto). Originator of several first-in-class drugs (first oral contraceptive, Celebrex co-development with Pfizer).
• Product embodying the claims? No. Endostatin was never commercialized by Searle, Pharmacia, or Pfizer as a marketed therapeutic. The claims cover a manufacturing/refolding process for a research-stage anti-angiogenic protein; the specification itself frames the work as enabling "preclinical studies and clinical trials," and I found no approved endostatin product from this chain. So the original assignee is an operating pharma company that did not ship a product reading on claim 1.
• Current status: Operating — but only as a wholly-owned subsidiary within Pfizer. Ownership chain of the corporate parent:
  • 1985 – G.D. Searle & Co. became the pharmaceutical unit of Monsanto Company.
  • April 2000 – Pharmacia & Upjohn merged with Monsanto (largely to acquire the Searle drug unit), forming Pharmacia Corporation; Monsanto's ag business later spun back out (2002).
  • April 16, 2003 – Pfizer Inc. acquired Pharmacia; Searle became a Pfizer subsidiary (G.D. Searle LLC).
• Bankruptcy: none in this chain. Monsanto, Pharmacia, and Pfizer were all solvent throughout; the Searle assets moved by merger and acquisition, not by insolvency sale.

Assignment timeline

Only one assignment for this patent is visible in the Google Patents legal-events record. It is the original inventor-to-company assignment. I could not retrieve its reel/frame or correspondent.

• 1999-01-26 (recorded) / execution date [not retrieved] — Reel [not retrieved] / Frame [not retrieved]
  • Conveyance: Assignment (assignment of assignors' interest / "ASSIGNMENT OF ASSIGNORS INTEREST — SEE DOCUMENT FOR DETAILS")
  • Assignor: Casperson, Gerald F.; Polazzi, Joseph O.; Violand, Bernard N. (note: Harding not listed)
  • Assignee: G.D. Searle & Co.
  • Correspondent: [not retrieved] — cannot assess recurrence across other tracked patents.
  • Context: Ordinary employment / inventor-to-employer assignment at filing time (internal capture of invention rights by the operating company), not an acquisition or reorg.

Subsequent corporate successorship — NOT reflected as recorded assignments in the sources reached:
The Monsanto → Pharmacia (2000) → Pfizer (2003) changes of control are corporate mergers, which under 35 U.S.C. § 261 / 37 CFR 3.21 may be recorded as "Merger" conveyances, and often are not separately recorded against every individual patent. I found no recorded post-issuance assignment for 6,653,098 (no change-of-name, no merger, no transfer-to-asserter). If the Assignment Center confirms none exists, the practical meaning is: the patent is still held within the original corporate family (now the Pfizer/G.D. Searle LLC side) and was never sold out of it.

Per your instructions: where the Assignment Center has no records beyond the original assignment, I say so plainly. Only the January 1999 original assignment is visible. I was unable to reach the Assignment Center to verify whether even a merger/change-of-name record exists beyond it — that specific negative is unresolved, so treat "only one assignment" as best available, not certified.

Timeline diagram

timeline
    title Ownership of US 6653098
    1998 : Provisional filed Feb 23
    1999 : Utility app filed Jan 14
         : Assigned to GD Searle and Co
    2000 : Monsanto merges into Pharmacia
    2003 : Patent issued Nov 25
         : Pfizer acquires Pharmacia
    2019 : Expired for fee reasons

NPE / troll-pattern signals

1. Shell-entity transfer — not present. The only recorded assignee is G.D. Searle & Co., an operating pharmaceutical manufacturer (recorded 1999-01-26). No "IP/Holdings/Ventures" LLC appears. Current listed assignee, GD Searle LLC, is a Pfizer operating subsidiary, not a single-purpose licensing shell. No registered-agent-service address is evident in the retrieved record.
2. Known asserter in the chain — not present. No assignee matches the Acacia / Marathon / IV / Wi-LAN-Conversant / Pendrell / Round Rock / Spangenberg NPE lists. Assignee set is Searle → (Monsanto) → Pharmacia → Pfizer.
3. Repeat correspondent across the chain — unclear (data gap). Correspondent of record for the 1999-01-26 recording was not retrieved, and there is only one visible link, so no recurrence can be assessed. A single appearance would not be a finding in any event.
4. Cascading transfers — not present. No chain of consecutive LLC-to-LLC assignments; no transfers at all beyond the original.
5. Pre-litigation transfer — not present. No litigation exists against this patent (consistent with the prior litigation search), so there is no assignment timed to precede a suit. The only transfer (1999) predates any conceivable assertion by 24+ years.
6. Bankruptcy fire-sale — not present. No Chapter 7/11 sale; the corporate history is merger/acquisition only (Pharmacia–Monsanto 2000; Pfizer–Pharmacia 2003).
7. Privateering — not present. No transfer of the patent from Pfizer/Pharmacia/Searle to a third-party NPE; the family stayed in-house.
8. Defensive aggregator (anti-NPE) — not present. Chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN.

Aggravating/atypical note (not an NPE signal): the inventor-vs-assignor mismatch (Harding named as inventor but absent from the assignor list; Casperson/Polazzi assigning but not named as inventors). This is a standing/chain-of-title housekeeping curiosity, not evidence of an NPE. It would only matter if the patent were ever asserted, which it was not.

Verdict

Insufficient data — the record shows only the original assignment (1999-01-26, assignors Casperson/Polazzi/Violand → G.D. Searle & Co.; reel/frame not retrieved), so under your rubric this lands squarely in the "only the original assignment" category.

Substance behind that formal verdict: every retrieved indicator points to a benign, operating-company-held, never-asserted patent — original assignee is a real pharmaceutical manufacturer, the chain stayed inside the Monsanto/Pharmacia/Pfizer corporate family after 2000–2003 mergers, there is no recorded secondary transfer, and there is no litigation. I could not confirm the reel/frame, correspondent, or execution date, so I will not assert an affirmative "clean" finding beyond what is documented.

Authoritative verification links:

• USPTO Assignment Center search for 6653098: https://assignment.uspto.gov/patent/index.html (search patent number 6653098)
• Google Patents legal events: https://patents.google.com/patent/US6653098/en
• USPTO Patent Center (ownership tab): https://patentcenter.uspto.gov (application 09/231,077)

Recommended follow-ups I could not complete: (a) pull the reel/frame + correspondent for the 1999-01-26 recording; (b) confirm whether a Merger conveyance (Monsanto→Pharmacia 2000, Pharmacia→Pfizer 2003) was ever recorded against this patent; (c) check Unified Patents' and RPX's asserter directories for "G.D. Searle LLC" / "6653098" to confirm the negative on assertion.

I'll search for each cited reference to build precise, grounded prior-art characterizations. Let me run several targeted searches.

Prior-Art Analysis for US Patent 6,653,098 B1 ("Method of producing mouse and human endostatin")

Scope note / compliance. Per the strict-ID rule I treated the number literally: 6653098 = US 6,653,098 B1. I did not have a tool that queries USPTO Patent Center directly; I worked from the authoritative full text of the patent (the USPTO document as mirrored, already retrieved in the prior section) and searched each of the six documents the patent itself cites. I did not substitute results for similar numbers (e.g., JP 6653098 — a sulfur-containing-polymer patent with coincidentally matching digits — was excluded). One housekeeping inconsistency: the task header says "Current Date: April 26, 2026" while the session date is 2026-09-30; nothing in this analysis turns on that difference.

This section builds on the earlier bibliographic/claim overview and does not repeat it.

1. Legal framework applied (pre-AIA § 102)

US 09/231,077 was filed 1999-01-14, claiming priority to provisional 60/075,587 (1998-02-23) — i.e., pre-AIA § 102 governs. Relevant critical dates:

• § 102(b) bar: publication more than one year before the U.S. filing date → before 1998-01-14.
• § 102(a): patented/described in a printed publication before the applicant's invention (≲ 1998-02-23).
• § 102(e): U.S. patent granted on an application filed in the U.S. before the applicant's invention date (for subject matter adequately supported).

Anticipation under § 102 requires every element of a claim in a single reference; a reference that supplies only some limitations is not an anticipatory reference (it may instead support § 103).

2. The cited references (as listed on the face of US 6,653,098)

The patent lists exactly six cited documents. Two of them (WO 97/15666 and US 5,854,205) are the same family (Children's Medical Center / O'Reilly–Folkman endostatin).

3. Per-reference analysis

Reference 1 — EP 0 312 358 A2

• Full citation: "Method for promoting intramolecular disulfide bond formation in recombinant proteins contained in a denaturant solution." Applicant: Mallinckrodt Veterinary, Inc. Priority 1987-10-14; published 1989-04-19. (Family member AU 615485 B2.)
• Description: Reduced recombinant protein in a denaturant solution (SDS, urea, or guanidine·HCl) is reacted with cystine to oxidize free sulfhydryls and form intramolecular (native) disulfide bonds. Working example: 0.1–2 % SDS, pH ~7–11, 0.1–3.0 mg/mL reduced recombinant somatotropin (rST), 0.5–10 mol cystine/mol protein, 12–36 h. Goal is to suppress incorrect inter-/intra-molecular bonds.
• § 102 analysis: Does not anticipate any claim. It is not an endostatin reference; the protein is somatotropin. It discloses cysteine/cystine redox chemistry in a denaturant and a broad pH range overlapping claim 1's 9–11.5, but it does not disclose (a) an endostatin-encoding polynucleotide in a prokaryotic host, (b) inclusion bodies of endostatin, (c) recovery of those inclusion bodies, the two-stage 4–10 M → 2–4 M urea regime, or the recitation of DTT/cysteine at 0.5–2 / 0.5–4 mM during refolding at 0.2–2 mg/mL. It is, at most, § 103-type art on disulfide-exchange technique (note claim 1 itself recites cystine nowhere; cystine appears only in the specification and dependent claims 11–13).

Reference 2 — EP 0 393 725 A1

• Full citation: "Process for activating genetically-engineered, heterologous eucaryotic proteins containing disulphide bridges after their expression in procaryotes." Applicant: Roche Diagnostics GmbH (parent: F. Hoffmann-La Roche). Priority 1985-10-22/23; published 1990-10-24 (grant 1990-10-23). Inventors Rainer, Mattes, Fischer. (Related family: EP 0 219 874 B1, CA 1 329 157 C, US 5,453,363.)
• Description: Disintegration of cells → solubilization under denaturing + reducing conditions → activation under oxidizing conditions in the presence of GSH/GSSG. Two stated variants: (i) activation at pH 9–12, GSH 0.1–20 mmol/L, GSSG 0.01–3 mmol/L, non-denaturing denaturant concentration; or (ii) separate reduction/denaturation (GSSG added under denaturing conditions to form protein–glutathione mixed disulfides), then activation at pH 7–10.5, GSH 0.5–5 mmol/L. Exemplified for t-PA and interferon-β.
• § 102 analysis: Does not anticipate any claim. It discloses high-pH solubilization and neutral-to-alkaline refolding of a disulfide-containing recombinant protein, and its pH windows overlap claims 1–5. But (i) the redox pair is GSH/GSSG, not the claimed DTT or cysteine at 0.5–2 / 0.5–4 mM; (ii) the protein is t-PA/interferon-β, not endostatin; (iii) no 4–10 M / 2–4 M urea two-stage urea regime is taught as claimed; (iv) no endostatin inclusion-body recovery. Only partial overlap → § 103-relevant, not anticipatory.

Reference 3 — WO 96/40736 A1

• Full citation: "Method of producing IGF-I and IGFBP-3 with correct folding and disulfide bonding." Applicant: Celtrix Pharmaceuticals, Inc. Priority 1995-06-07 (US 08/482,271, which issued as US 5,789,547); filed 1996-05-30; published 1996-12-19. (EP 0 832 098 B1; DK 0832098 T3; ES 2179943 T3; CN 1214689 A.)
• Description: Cofolding of IGF-I together with IGFBP-3 in a single reaction — denature + reduce IGFBP-3, add native IGF-I, then add oxidizing agent — to raise yields of correctly folded IGF-I/IGFBP-3 complex.
• § 102 analysis: Does not anticipate any claim. Different protein system (IGF axis), different strategy (cofolding a binding partner), no endostatin, no two-stage urea concentrations, no claimed reductant range. Not anticipatory for any of claims 1–26.

Reference 4 — WO 97/15666 A1 (the closest endostatin art)

• Full citation: "Therapeutic antiangiogenic compositions and methods." Applicant: The Children's Medical Center Corporation (inventors O'Reilly, M.S.; Folkman, M.J.). Priority 1995-10-23; PCT-US filing published 1997-05-01.
• Description: The foundational endostatin filing. Discloses endostatin (~20 kDa C-terminal fragment of collagen XVIII; SEQ of the N-terminus HTHQDFQPVLHLVALNTPLS…), purification from EOMA-conditioned medium, and recombinant expression in E. coli and in baculovirus-infected insect cells. Critically for this patent, it discloses mouse and human endostatin, E. coli expression yielding an insoluble inclusion-body-derived material, and its use as an insoluble suspension — and expressly reports that a refolding attempt (0.1 M Na-phosphate pH 7.4, 150 mM NaCl, 0.6 M urea, 2 mM GSH, 0.02 mM GSSG, 0.5 M arginine; 0.1 mg/mL) lost >99 % of the protein, which is why the unrefolded suspension was used for in vivo work.
• § 102 analysis: Does not anticipate claim 1 (or any claim). It discloses elements (a)–(c)-type subject matter — a prokaryotic host expressing endostatin that forms inclusion bodies — but it does not disclose element (e), "properly refolding the endostatin at about neutral pH" under the claimed urea/reductant regime, nor element (f) "isolating properly folded, active forms." The one refolding experiment described uses 0.6 M urea and a GSH/GSSG redox pair at 0.1 mg/mL, none of which matches claim 1's 2–4 M urea / DTT 0.5–2 mM or cysteine 0.5–4 mM / 0.2–2 mg/mL — and it is taught as a failure, i.e., it teaches away from the claimed solution rather than anticipating it.
• Caveat: WO 97/15666 would be anticipatory of a claim drawn only to "an isolated polynucleotide encoding endostatin" or "a method of expressing endostatin in E. coli to form inclusion bodies" — but no such claim exists; every claim 1–26 requires the refolding/isolation steps and/or the specific numeric parameters.

Reference 5 — WO 97/18233 A1 (the second-closest art, but different protein and a contrary process step)

• Full citation: "Method for producing a correctly folded, biologically active recombinant protein." Applicant: Pharmacia & Upjohn AB. Priority 1995-11-13; published 1997-05-22. (DE 696 31 906 T2 family.)
• Description: Steps: (a) express the protein/polypeptide in prokaryotic cells (preferably E. coli); (b) harvest the cells; (c) "directly solubilization of the cells in a buffer at pH about 8 to 11 … with a chaotropic agent and a reducing agent"; (d) dilute with water and a diluent (ethanol, optionally guanidine·HCl). The chaotrope is preferably guanidine (3–7 M, pref. 5 M) or urea; the reductant is DTT or cysteine. Target proteins: IGF-I, IGF-II, GH (preferably a Z-IGF-I fusion). The specification emphasizes the process is conducted "in the absence of mechanical disruption of the cells and isolation and washing of refractile bodies."
• § 102 analysis: Does not anticipate claim 1. Two independent reasons:
  1. Different analyte — the method is directed to IGF-I/IGF-II/GH, not endostatin.
  2. Element (c) conflict — claim 1 requires "recovering said inclusion bodies," whereas WO 97/18233 explicitly minimizes/avoids isolating refractile bodies by solubilizing whole cells directly. It therefore teaches away from claim element (c) rather than disclosing it. Its pH 8–11 range and DTT/urea disclosure overlap claim limitations, but the reference cannot supply the endostatin + inclusion-body-recovery combination. § 103-relevant only.

Reference 6 — US 5,854,205 A

• Full citation: "Therapeutic antiangiogenic compositions and methods." Assignee: The Children's Medical Center Corporation (O'Reilly, Folkman). Priority/filed 1995-10-23; issued 1998-12-29.
• Description: U.S. counterpart of WO 97/15666 (Reference 4) — endostatin compositions and nucleic acids; same E. coli / inclusion-body-expression and unrefolded-suspension disclosure. (Later same-family grants include US 6,346,510 and US 6,764,995.)
• § 102 analysis: Does not anticipate any claim — same reasoning as Reference 4. Because it issued after the 1998-01-14 § 102(b) critical date, it is not a § 102(b) reference; it is potentially § 102(e) art (U.S. patent granted on an application filed 1995-10-23, before the applicant's 1998-02-23 invention date) but only for subject matter it supports — and that subject matter still lacks the claimed refolding/isolation steps and numeric parameters.

4. Bottom line on § 102

None of the six cited references anticipates any of claims 1–26 of US 6,653,098 under 35 U.S.C. § 102.

• The only endostatin references among the citations (WO 97/15666 and its U.S. counterpart US 5,854,205) supply, at most, elements (a)–(c) but not the claimed high-pH-urea-solubilization / neutral-pH-urea-refolding / DTT-or-cysteine redox / 0.2–2 mg/mL isolation combination of claim 1 — and they affirmatively report that refolding as attempted failed (>99 % loss), which is a teaching away, not a disclosure of the claim.
• The generic protein-refolding references (EP 0 312 358, EP 0 393 725, WO 96/40736, WO 97/18233) supply only scattered, overlapping limitations (pH windows, cystine/DTT/cysteine, chaotrope choice) applied to other proteins (somatotropin, t-PA, interferon-β, IGF-I/IGFBP-3); each is missing multiple claim-1 elements, and WO 97/18233 in particular contradicts the "recovering said inclusion bodies" limitation.

Practical consequence: these citations are best characterized as § 103 (obviousness) art, not § 102 art. The likely examiner rationale for allowance (consistent with the patent's own narrative) is that the cited art showed endostatin could be expressed but not productively refolded, and showed generic refolding methods for unrelated proteins — so the specific, working endostatin refolding regime was non-obvious and unsupported by any single anticipatory reference.

5. Two grounding caveats

1. Post-priority documents are not § 102 art here. The "Cited By" and "Families Citing" entries in the record (e.g., Biogen Idec, Bolder Biotechnology, Amgen refolding filings; Merck's and the U.S. Government's later endostatin applications) post-date the 1998 priority date and therefore are not available as § 102 prior art against this patent. They are relevant only as evidence of how the field evolved. Likewise RU 2278688 C1 and similar foreign documents merely cite US 6,653,098.
2. Most directly on-point art is non-patent literature, not in the citation list. The references the patent discusses — O'Reilly et al., Cell 88:277–285 (1997); Boehm et al., Nature 390:404–407 (1997); Standker et al., FEBS Lett. 420:129–133 (1997) — are the true technological backdrop. Under § 102 they fare no better than the patent citations: O'Reilly 1997 discloses E. coli endostatin as an insoluble suspension and reports the failed refolding, i.e., it too fails to disclose element (e)/(f) of claim 1.

If a definitive negative-anticipation confirmation is needed, the authoritative checks are the USPTO Patent Center / Global Dossier "References Cited" and "Continuity" tabs and the examiner's citation listing in the file wrapper (Form PTO-892), which would confirm whether any additional § 102 art was applied during prosecution beyond the six documents reproduced in the published record.

I'll ground the prior-art analysis by pulling details on the cited references before constructing the §103 combinations.

Let me pull details on the remaining cited references.

Obviousness Analysis — US 6,653,098 B1 under 35 U.S.C. § 103

Note on the record I'm working from. This analysis is built on the "Citations (6)" set listed on the face of US 6,653,098 and on the patent text already summarized in the prior sections. I could not retrieve the actual prosecution history (office actions, rejections, applicant remarks), so I cannot say which references the examiner actually applied or what was overcome. The analysis below is a de novo §103 assessment, not a reconstruction of the file history.

Small metadata flags. (1) The task header gives the current date as April 26, 2026, while the system context gives September 30, 2026 — I note the inconsistency but it does not affect the analysis. (2) As flagged in the earlier sections, SEQ ID NOs 10 and 11 are both machine-tagged "PRT / human" in the sequence listing although the specification defines SEQ ID 10 as mouse endostatin. I do not correct this; I treat the specification text as controlling for claim 22 vs. claim 24 purposes.

1. The § 103 framework

Under Graham v. John Deere Co., 383 U.S. 1 (1966), the analysis turns on (1) the scope and content of the prior art, (2) the differences between the prior art and the claims, (3) the level of ordinary skill, and (4) objective evidence of nonobviousness. Under KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007), a claim is obvious where the prior-art elements are known and there was a reason to combine them — including "common sense," the finite number of identified, predictable solutions, and the fact that the field is one of "ordinary creativity" rather than unpredictable art. Protein refolding from E. coli inclusion bodies in 1998 was exactly such a mature, parameter-driven, predictable field: the patent's own Background section concedes this, citing Rudolph (Protein Engineering: Principles and Practice, 1996) and Fisher, Sumner & Goodenough (Biotechnol. Bioeng. 1:3-13, 1992) for the general body of solubilization/refolding practice.

Level of ordinary skill (PHOSITA)

A person of ordinary skill as of the Feb. 23, 1998 priority date would be a scientist with a Ph.D. in biochemistry, molecular biology, or biochemical engineering (or a B.S./M.S. with several years of hands-on experience) who routinely: designs E. coli expression constructs; isolates inclusion bodies by homogenization/centrifugation; solubilizes them in 4-8 M urea or 4-6 M guanidine-HCl with a reductant; and refolds by dilution into redox-buffered, low-denaturant, dilute solutions. That profile is confirmed by the level of the cited art itself — EP 0393725, EP 0312358, and WO 97/18233 each assume the reader already knows inclusion-body processing as a matter of course.

2. Scope and content of the prior art (the six references)

Key structural observation: these six references are not scattered across fields. They are all directed to the same problem — recovering correctly folded, disulfide-bonded recombinant protein from prokaryotic inclusion bodies — and four of them (EP 0393725, EP 0312358, WO 96/40736, WO 97/18233) expressly address the same sub-problem: forming the correct disulfide pairings during renaturation. Under KSR, that shared field is itself a powerful motivation to combine.

3. Claim 1 element-by-element chart

4. The obviousness combinations

Combination A (primary): WO 97/15666 / US 5,854,205 + EP 0393725

The combination. WO 97/15666 supplies every endostatin-specific element — the gene, the E. coli host, inclusion-body formation, the sequences of SEQ ID NOs 5-9 and 10-11, and the express recognition that (i) the E. coli material must be refolded and (ii) the art's one prior attempt failed. EP 0393725 supplies the missing generalizable process: solubilize the refractile bodies under denaturing and reducing conditions, then reactivate under oxidizing conditions at alkaline-to-neutral pH with a non-denaturing residual concentration of denaturant (expressly urea 0.5-4 M, pref. 1-3.5 M) in the presence of a GSH/GSSG redox couple.

Why a PHOSITA would combine them. (1) Same field and same problem — both concern activating disulfide-bridged recombinant protein recovered as insoluble aggregates from prokaryotes. (2) The endostatin reference states the need: WO 97/15666 documents that >99% of protein was lost in its refolding attempt and that the authors resorted to injecting an insoluble suspension — a result it characterizes as precluding in vivo assays. That is an express invitation to the art to supply a better renaturation protocol. (3) No teaching away: EP 0393725 criticizes only copper-ion catalysis and single-step dilution into strong denaturant; it does not disparage alkaline solubilization or low-denaturant refolding, and it is not limited to t-PA/IFN-β — its claim 1 is generic to "disulfide bridge-containing eukaryotic proteins." (4) Predictable result: the claimed output — soluble, correctly disulfide-paired, active protein — is precisely what EP 0393725 was designed to produce.

Overlap of ranges (a strong, independent basis). Under In re Peterson, 315 F.3d 1325 (Fed. Cir. 2003), and In re Wertheim, 541 F.2d 257 (CCPA 1976), where a claimed numerical range overlaps or is contiguous with a range disclosed in the prior art, a prima facie case of obviousness exists unless the applicant shows criticality. Here:

• Claimed solubilization pH 9-11.5 directly overlaps Pharmacia's pH 8-11 and Roche's pH 8-12 / 9.5-11.
• Claimed refold urea 2-4 M directly overlaps Roche's urea 0.5-4 M (pref. 1-3.5 M).
• Claimed refold protein concentration 0.2-2 mg/mL is within Mallinckrodt's disclosed 0.1-3.0 mg/mL.

The only claim-1 value that is not numerically pre-figured is the juxtaposition: elevated pH at solubilization and neutral pH at refolding. But the specification does not allege that this ordering is critical — it says only that "other pH conditions… can be used, but at lower efficiency." That concession is fatal: it frames the pH regime as a yield-optimization knob, which is the paradigm of routine experimentation (In re Aller, 220 F.2d 454 (CCPA 1955); In re Applied Materials, 692 F.3d 1289 (Fed. Cir. 2012)). Moreover, the reason alkaline pH helps solubilization (increased protein solubility at alkaline pH, favored ionization of acidic side chains) is a general, well-understood protein-chemistry principle, and WO 97/18233 had already applied it to inclusion-body solubilization.

Combination B: WO 97/15666 + EP 0312358 (cystine-mediated disulfide formation)

Although claim 1 does not itself recite an oxidant/disulfide-interchange reagent, the specification's operative mouse-endostatin example adds cystine to 10 mM immediately before dilution, and dependent claims 14-17 claim disulfide formation by air oxidation over 12-96 h (pref. ~60 h). EP 0312358 discloses exactly this: reacting a reduced recombinant protein in a urea denaturant solution with cystine to promote intramolecular disulfide formation, over ~12-36 h, at a protein concentration of 0.1-3.0 mg/mL and pH ~7-11, and reports 50-70% native pairing with reduced aggregates. A PHOSITA would add cystine to the refold for the reason EP 0312358 gives — avoiding the mispaired inter- and intra-molecular disulfides that copper catalysis produces — and would arrive at the ~60 h air-oxidation dwell by routine monitoring (as the patent itself did, "as assessed by reverse-phase HPLC"). Notably, this is also why claim 1's requirement that a reductant (DTT/cysteine) be present during refolding is not inventive: the controlled-reductant/low-oxidant-donor regime is the classic redox-buffer approach of EP 0393725 and WO 96/40736.

Combination C: WO 97/15666 + WO 97/18233 (Pharmacia)

WO 97/18233 independently teaches the "elevated-pH solubilization → dilute to fold" architecture in a prokaryotic host: solubilizing directly in a buffer at pH ~8 to 11 with a chaotropic agent and a reducing agent, followed by dilution to obtain correctly folded, biologically active protein. Combined with WO 97/15666's endostatin genes and inclusion bodies, this renders the (d)/(e) pH architecture obvious without Roche. The claimed 0.2-2 mg/mL refold concentration is likewise the natural consequence of "diluting" per WO 97/18233.

Combination D: WO 97/15666 + WO 96/40736 (Celtrix)

WO 96/40736 teaches an explicit redox-controlled refolding sequence — denaturation and reduction in a denaturant plus reducing agent, then oxidation to form native disulfides, with discussion of disulfide-exchange equilibration. Substituting a single protein (endostatin) for the IGF-I/IGFBP-3 pair is an obvious substitution of one known recombinant disulfide-containing protein for another, with an expectation of the same class of result (In re Kemps, 97 F.2d 1427 (CCPA 1938); In re Fracalossi, 1999 WL 33117388).

Combination E: any of A-D + the admitted general knowledge

The claimed solubilization conditions (urea 4-10 M; ~6 M preferred) and the solubilization diluent hold time, the choice between DTT and cysteine as reductant, and the reducing-agent concentrations (2-10 mM solubilization; 0.5-2 mM DTT / 0.5-4 mM cysteine refold) are all within the routine-optimization zone expressly acknowledged in the patent's Background (6-8 M urea; 6 M GuHCl) and catalogued in the standard review literature the patent itself cites. No unexpected property is asserted for any single parameter.

5. The dependent claims

Because claims 2-26 all depend from claim 1, they rise or fall with it — and each adds only a narrower parameter value in the same, already-disclosed range:

• Claims 2-3 (pH 10-11, ~10.5) — within Roche's pref. 9.5-11; optimization.
• Claims 4-5 (refold pH 7-8, ~7.5) — the physiological/neutral window; matches WO 97/15666's pH 7.4 attempt.
• Claims 6-7 (urea 6 M solubilization; 3.5 M refold) — 3.5 M is literally Roche's preferred upper end ("pref. 1-3.5 M"); 6 M is the admitted standard.
• Claims 8-13 (DTT/cysteine concentrations) — reductant titrations; routine.
• Claims 14-17 (air oxidation 12-96 h; 24-72 h; ~60 h) — EP 0312358's 12-36 h plus routine extension.
• Claims 18-20 (1-20 mg/mL solubilization, 2.5 mg/mL; 0.25 mg/mL refold) — within EP 0312358's 0.1-3.0 mg/mL (refold) and routine solubilization loadings.
• Claim 21 (ion-exchange, HIC, RP-HPLC purification) — WO 97/15666 and EP 0393725 both employ such steps; the specification itself cites Methods in Enzymology vol. 182.
• Claims 22-25 (mouse endostatin; SEQ ID NOs 5-8; human endostatin; SEQ ID NO 9) — WO 97/15666 (and US 5,854,205) disclose the mouse endostatin and human collagen XVIII C-terminal sequences. Reciting a known nucleotide sequence as the starting material does not impart patentability to an otherwise obvious process.
• Claim 26 (E. coli) — the express host of WO 97/15666 and EP 0393725.

6. Objective evidence of nonobviousness — assessment

The strongest available Graham factor for the patentee would be unexpected results / failure of others, and it deserves a fair hearing:

• Failure of others — WO 97/15666 reports that its refolding trial lost >99% of the protein, and that the authors "used the uncharacterized insoluble (nonrefolded) form of endostatin for most of their in vivo studies." The patent's own text repeats this. That is real evidence that practitioners had not yet solved endostatin refolding.
• Long-felt need — the specification's closing paragraph describes the commercial scale-up problem (soluble, well-characterized material for preclinical/clinical use).

But neither is decisive here, for three reasons:

1. The failure is attributable to conditions outside the claim, not to the claimed solution. WO 97/15666's failure came from refolding at 0.6 M urea with GSH/GSSG/arginine at 0.1 mg/mL — deliberately outside the claimed 2-4 M urea refold window. A single failed attempt at one parameter set does not teach away from the field, and KSR cautions that "the fact that a particular combination was tried before and failed may be relevant" but does not convert routine optimization into invention. EP 0393725 had already demonstrated that alkaline-stage, low-denaturant reactivation with a controlled redox couple works for E. coli-derived disulfide-bridged proteins.
2. No criticality data. The specification never compares its yields to a Roche-style or Mallikckrodt-style protocol; it reports only RP-HPLC traces (FIGS. 3-6) at varying pH and urea and cell-based activity assays (FIGS. 7-10). It asserts improved quality of material, not a statistically surprising yield, and it concedes that other pH/urea conditions "can be used, but at lower efficiency" — the language of a response surface, not of a criticality discovery.
3. Weak commercial-signal evidence. The patent went Expired – Fee Related (anticipated expiration 2019-01-14), and no litigation or licensing of record was found (per the earlier Litigation section). There is no affirmative evidence of commercial success, copying, industry praise, or licenses attributable to the claimed subject matter.

Counter-consideration I should flag honestly: if the applicant could show, in the file history or in declarations, that the specific combination of elevated-pH solubilization with 2-4 M urea at neutral pH produced yields an order of magnitude above what the art's teachings would predict — i.e., true criticality — the §103 conclusion could flip for claim 1. That evidence is not in the patent as issued and I have not seen it.

7. Conclusion

Claim 1 is prima facie obvious over WO 97/15666 (or its U.S. sibling US 5,854,205) in view of EP 0393725 — and, in the alternative, over WO 97/15666 in view of WO 97/18233, EP 0312358, or WO 96/40736. The motivation is supplied by the references' shared field, WO 97/15666's own articulation of the unmet need, and the finite, well-catalogued set of refolding parameters. The numerical limitations are either disclosed outright (pH, urea, protein concentration ranges overlap the art) or are routine optimizations of known variables. The dependent claims add only narrower values within those disclosed ranges, together with a known E. coli host and known endostatin sequences.

Anticipation (§102): no single one of the six references discloses all of claim 1. WO 97/15666 lacks the elevated-pH solubilization and the claimed 2-4 M urea refold; EP 0393725 is not directed to endostatin; neither WO 97/18233 nor EP 0312358 discloses endostatin. So the case for invalidity rests on §103, not §102 — but on §103 it rests on multiple independent, mutually reinforcing combinations rather than on a single precarious one.

Caveats. (i) I have not seen the prosecution history, so I cannot identify what the examiner actually relied on or what the applicant argued. (ii) The ultimate legal conclusion requires the full intrinsic record and any Rule 132/§103(c) submissions. (iii) The "prior art" universe used here is the six-reference citation set from the patent record; a full invalidity analysis would also sweep non-patent literature (e.g., Rudolph & Lilie; Fischer et al., Biotechnol. Bioeng. 1:3-13, 1992; Creighton, Methods Enzymol. 107:305-329) already cited inside the patent and admitted as background.