US 5658759

I'll verify the patent against live sources before summarizing.

US 5,658,759 — verified summary

Caveat on scope of verification: the details below are drawn from the patent's own full text plus live web results (Google Patents, FreePatentsOnline, docket records). I searched for the literal number 5658759 and 5,658,759 and did not find any CAFC 2026 docket entry involving this patent. My search was via general web search rather than a direct USPTO PatentCenter/CAFC PACER query, so treat "no 2026 CAFC activity found" as a negative search result, not an authoritative absence.

Bibliographic data

Assignee caveat: Google Patents lists both "Lonza Group AG" and "Rio Tinto Switzerland AG" as current assignees and expressly disclaims accuracy of assignee listings. The recorded assignment chain on that page runs Celltech Limited → Alusuisse Holdings A.G. (1997-11-10) → Alusuisse Lonza Group AG (merger, 2001-07-23) → Lonza Group AG (2001-07-23). I could not independently verify a "Rio Tinto" record, and I flag it as an unverified data artifact rather than asserting it.

Abstract (as issued)

The invention provides expression vectors containing the promoter, enhancer and substantially complete 5'-untranslated region including the first intron of the major immediate early gene of human cytomegalovirus. Further vectors including the hCMV-MIE DNA linked directly to the coding sequence of a heterologous gene are described. Host cells transfected with the vectors and a process for producing heterologous polypeptides using the vectors and the use of the hCMV-MIE DNA for expression of a heterologous gene are also included within the invention.

Plain-language overview of the claims

The claim set has one independent claim (claim 1); claims 2–10 are all dependent.

• Claim 1 (independent) — the host cell. A cell that has been transfected with a recombinant expression vector, where the vector carries three hCMV-MIE elements — the promoter, the enhancer, and the complete 5' untranslated region including the first intron — hooked up ("operably linked") to a foreign (heterologous) protein-coding sequence. Plainly: a cell engineered to make a chosen foreign protein under the control of the full hCMV major immediate early regulatory/leader region.

• Claim 2 — manufacturing method (broad). A process for recombinantly producing a heterologous polypeptide by culturing the claim-1 host cell. This is the workhorse production claim.

• Claim 3 — convenience restriction site. The claim-1 cell where the vector also has a restriction site to make it easy to slot the heterologous coding sequence in.

• Claim 4 — direct fusion. The claim-1 cell where the hCMV-MIE promoter/enhancer/complete 5'-UTR (with first intron) is joined directly to the coding sequence — no intervening 5'-UTR of the target gene. This reflects the specification's stated discovery that direct linkage gave consistently high translation.

• Claim 5 — native translation start included. The claim-1 cell where the vector also retains the hCMV-MIE translational initiation signal.

• Claim 6 — specific plasmids. The claim-1 cell where the vector is one of the named constructs: pCMGS, pEE6.hCMV, or pHT.1.

• Claims 7, 8, 9, 10 — production methods narrowed. Processes for recombinantly producing a heterologous polypeptide by culturing the host cells of claims 3, 4, 5, and 6 respectively. These are the same culturing step as claim 2, but tied to the narrower vector configurations.

Technical context (from the specification)

The disclosed constructs use the Pst-I fragment of hCMV (citing Boshart et al., Cell 41:521-530 (1985) and Spaete & Mocarski, J. Virol. 56(1):135-143 (1985)), which contains the promoter-enhancer and most of the 5'-untranslated leader including the first intron; the remainder of the 5'-UTR is recreated with a ~20 bp synthetic oligomer that also introduces an NcoI site (5'-CCATGG-3') at the initiation codon, improving Kozak consensus. Worked examples report TIMP secretion from a single-copy CHO-K1 line at 3×10⁸ molecules/cell/day, rising to 3×10⁹ molecules/cell/day after one round of MSX amplification (clone GS-19.6(500)14), plus PA-1 and PA-2 plasminogen activator secretion, GS/MSX selection, and transient light-chain expression in COS-1 cells.

Litigation of record (not 2026)

The asserted litigation I could locate is from 2007, not 2026:

• Lonza Group AG v. Northwest Biotherapeutics, Inc., 8:07-cv-03177 (D. Md., filed 2007-11-27) — asserted US 5,658,759 among eight patents (also 5,122,464, 5,591,639, 5,770,359, 5,827,739, 5,879,936, 5,891,693, 5,981,216) over DCVax products.
• Lonza Group AG v. Northwest Biotherapeutics, Inc., 1:07-cv-00467 (D. Del., filed 2007-07-27) — related filing listing the same patent family; the AO-120 form records 5,658,759 with holder "Celltech Limited."

Sources: Google Patents US5658759A, FreePatentsOnline 5658759, Docket listing 8:07-cv-03177, patentdocs report of the Lonza/Northwest complaints. No CAFC 2026 appeal docket for this patent was found in my searches.

I'll research litigation involving US 5,658,759 using the sources you specified.

Let me dig deeper on the specific case referenced in the patent record.

I found the key case. Let me verify the details.

I'll verify the canonical "no proceedings" default against live sources before reporting.

Let me probe a few more angles — reexamination history, the Lonza/Northwest dispute, and any CBM/PGR filings.

One more targeted check on reexamination and any post-issuance USPTO proceedings.

Proceedings overview

Total AIA trial proceedings on US 5,658,759: zero. The canonical USPTO Open Data Portal block attached to this prompt returns no IPR, PGR, or CBM proceedings, and my independent web searches (Google Patents litigation/PTAB field, FreePatentsOnline, PTAB decision pages, and general queries on "5,658,759" + IPR / PTAB / Lonza / Celltech / Unified Patents) surfaced no petition, no institution decision, no Final Written Decision, and no appeal — so the breakdown is 0 active / 0 claims invalidated / 0 claims sustained / 0 settled / 0 institution denied. All ten claims of 5,658,759 (claims 1–10 as issued) are therefore untested at the Board but expired: Google Patents records an anticipated expiration of 2009-04-28 with status "Expired – Fee Related," which predates the 2012-09-16 effective date of AIA trials by more than three years. The bottom line for a defendant is not "hardened patent" or "dead claims" — it is "expired patent with no PTAB record either way," meaning there is no FWD to cite and no § 315(e)(2) estoppel to exploit, but also no live term over which damages can accrue.

No proceedings to report

I want to be explicit rather than pad this section with invented dockets:

• No IPR. No petition number (IPR20xx-xxxxx) referencing patent 5,658,759 appears in any indexed source.
• No PGR. Unavailable in any event: PGR under § 321 applies only to patents with an effective filing date on or after 2013-03-16, and this patent's priority date is 1987-07-23.
• No CBM. The patent is directed to host cells and expression vectors (C12N15/85, C07K14/8146), not to a "financial product or service," so it was never CBM-eligible; the CBM window closed 2020-09-16 regardless.
• No ex parte or inter partes reexamination on this patent surfaced in my searches. (The 2005–2008 reexaminations I did find in this technology space concerned Genentech's Cabilly II patent, not 5,658,759 — see the Genentech answer filed 2010-03-24, which also references a 1998 § 146 action, Genentech v. Celltech, D. Del. No. 98-3926. That is a district-court interference appeal, not a Board proceeding.)

Verification is limited by tooling: the ODP API block is the authoritative negative, and my searches were general web queries rather than a direct PTAB E2E / PACER pull. Treat this as a well-supported negative search result, not a certifiable absence.

Nearest-neighbor matters (not PTAB proceedings on this patent — included only to close the loop)

Third-party reporting indicates Lonza withdrew nearly all claims in the Maryland action within about five months and the residual DCVax-Prostate claim resolved via inventory destruction, with no monetary payment and no license taken. I could not verify those settlement terms in a primary source — treat as secondary-source reporting, and note that settlements are typically confidential anyway.

Strategic summary

Claim status. No claim of 5,658,759 has ever been canceled, confirmed, or construed by the PTAB. Claims 1–10 (claim 1 being the sole independent claim — a host cell transfected with a vector carrying the hCMV-MIE promoter, enhancer, and complete 5'-UTR including the first intron, operably linked to a heterologous coding sequence) remain exactly as issued on the face of the patent. That is unusual for a patent that was heavily asserted in 2007, and the explanation is almost certainly temporal: the AIA trial regime did not exist during the patent's commercially relevant life, and by the time IPRs became routine the patent was already expired, making a petition economically pointless. Note the near-identical sibling US 5,591,639 (division parent, same title), whose claims are drawn to the vectors and plasmids (pCMGS, pEE6.hCMV, pHT.1) rather than host cells — also expired, also with no PTAB record I could find.

Estoppel landscape. There is none. Section 315(e)(2) estoppel attaches only after institution and an FWD, and neither occurred here. There is consequently no estoppel bar on any prior-art ground — a defendant could in principle raise anticipation or obviousness in district court based on the very references already of record (Boshart et al., Cell 41:521-530 (1985); Spaete & Mocarski, J. Virol. 56(1):135-143 (1985); Pasleau et al., Gene 38:227-232 (1985); Stenberg et al., J. Virol. 49(1):190-199 (1984); Thomsen et al., PNAS 81:659-663 (1984); Foecking & Hofstetter, Gene 45:101-105 (1986); EP 260148) free of § 325(d) implications at the Board, because no Board proceeding exists to which § 325(d) could apply. The practical point, though, is that this freedom is worthless to a defendant: there is no live infringement theory to defend against.

Pattern signals. No repeat petitioner, no defensive aggregator (no Unified Patents or RPX involvement appears in the chain), and no patent-owner PTAB appeal activity — consistent with a patent that was monetized through ordinary district-court assertion in 2007 and has been dead since roughly the end of the last decade. The Darts-IP "family has litigation" flag in the Google Patents record traces to the 2007 Lonza/Northwest filings, not to any post-grant proceeding.

Recommended next steps

1. Lead with expiration, not with supposed IPR wins. There is no FWD to quote and no canceled claim to point at. Do not let anyone on the team assert that the claims were invalidated — they were not, and saying so would be a factual misrepresentation. The correct sentence is: "US 5,658,759 expired and has never been the subject of an AIA trial; a complaint filed today cannot reach any pre-expiration infringement under 35 U.S.C. § 286."
2. The § 286 point is robust to the expiration-date uncertainty. The structured data lists anticipated expiration as 2009-04-28. Even if the term instead ran the longer post-URAA route to 2014-08-19 (17 years from the 1997-08-19 issue date, per 35 U.S.C. § 154(c)(1) for pre-URAA applications), § 286 bars recovery for infringement more than six years before suit — so a complaint filed on or after 2020-08-20 recovers nothing. As of today, 2026-09-27, the recovery window is empty either way. I flag that the 2009 date is quoted data, not a term computation I verified (terminal-disclaimer history was not checked).
3. If you want a clean Paper trail, pull the primary records yourself. Run the patent number through PTAB E2E and USPTO PTAB Decisions, and check CourtListener for any Federal Circuit appeal naming Bebbington, Celltech, Alusuisse, or Lonza with 5,658,759. My searches found no CAFC appeal of any PTAB decision on this patent, which follows mechanically from there being no PTAB decision.
4. If a demand letter arrives anyway, redirect the analysis to the sibling patents. A licensor asserting 5,658,759 in 2026 is likely asserting a family — check 5,591,639 (vectors/plasmids) and 5,981,216, 5,879,936, 5,891,693, 5,770,359, 5,827,739, 5,122,464, all of which appear in the same 2007 complaints. Those are likewise pre-1995 filings and expired, but confirm each one's expiration and maintenance-fee history rather than assuming symmetry across the family.
5. Document the CDMO-vendor exposure point. The 2007 Lonza theory was that a contract manufacturer's supplier used the patented expression system — i.e., exposure ran through a CDMO, not the sponsor's own construct. For any current manufacturing relationship using a GS/CHO expression platform with hCMV-MIE-derived regulatory elements, confirm chain-of-title and the supplier's own license posture rather than relying on the expiration of this one patent.

Caveats I am not papering over: (a) the absence of PTAB proceedings is a negative search result from the ODP block plus web indexing, not a certified clearance; (b) the "Rio Tinto Switzerland AG" assignee entry in the Google Patents record remains unverified and looks like a data artifact; (c) the Northwest settlement terms are secondary-source only.

I'll research the assignment record for US 5,658,759. Let me start with the primary sources.

Let me search for the specific reel/frame records and the anomaly in the chain.

Let me try to pull the specific reel/frame records and check the anomalous filing.

⚠️ Data-access caveat, stated up front (read this before relying on any reel/frame reference below). The USPTO Assignment Center (assignmentcenter.uspto.gov) and its legacy mirror (assignment.uspto.gov) are JavaScript applications whose result tables are not exposed to general web crawling. My only tool for this task is general web search, and repeated queries for the literal strings 5658759, "Alusuisse Holdings" + "reel" + "frame", and "5658759" AND "reel/frame" returned no reel/frame numbers, no correspondent names, and no execution dates for this patent. What I do have is the Google Patents "Legal Events" mirror of the assignment records (conveyance types match USPTO "SEE DOCUMENT FOR DETAILS" boilerplate verbatim), which is reproduced in the authoritative full text supplied to me. I therefore present the recorded conveyance chain below without reel/frame and without correspondent, and I have left those fields explicitly marked "not retrieved" rather than fabricating them. Any analyst needing the reel/frame must query the Assignment Center directly by patent number.

Inventors

Pattern notes:

• Single-inventor patent. No co-inventor departures to track. The "all inventors left within 12 months → fire-sale" pattern is not assessable on a one-inventor patent and I found no evidence either way as to when Bebbington left Celltech. I am not asserting a departure.
• The relevant "sale" in this family was not inventor-driven but business-unit driven: Celltech sold its Celltech Biologics contract-manufacturing/expression-technology business to the Swiss Alusuisse group in 1996, which is the transaction that produced the 1997 assignment below.

Original assignee

• Entity on the issued patent: Celltech Limited (Celltech Ltd), UK.
• Primary business: Europe's first biotechnology company (founded 1980); recombinant DNA, mammalian cell expression technology, and biopharmaceuticals. The patent's subject matter — the hCMV-MIE promoter/enhancer/leader driving a coding sequence, paired with a GS selection marker — is the core of what became the industry-standard "GS Gene Expression System."
• Did they ship a product embodying the claims? Yes, as a technology/process rather than a consumer product. The claims are directed to a transfected host cell and a method of culturing it; the commercial embodiment is the CHO/NS0 GS expression platform and cell lines built on it. Celltech's Biologics division supplied this platform under licence. The technology is recited as background in dozens of later patents (e.g., references to US 5,658,759 alongside 5,591,639 / 5,981,216 / WO 87/04462 as the GS-system description).
• Current status of original assignee: Acquired / no longer independent. Celltech's biologics business was sold to Alusuisse-Lonza (1996), becoming Lonza Biologics; the residual Celltech Chiroscience merged in 1999 and the group was acquired by UCB (Union Chimique Belge) in 2004. Celltech Limited no longer exists as an operating brand.

Assignment timeline

Reel/frame and correspondent: not retrieved. As explained in the caveat above, neither field was surfaced by my search access. The conveyor sequence, conveyance types and dates below are taken from the Google Patents legal-events mirror of the USPTO assignment records. Do not treat the absence of a reel number as no record existing — a record clearly exists; I simply could not read its identifier.

• Executed date not retrieved / recorded 1997-11-10 — Reel not retrieved / not retrieved

  • Conveyance: Assignment of assignors' interest ("see document for details")
  • Assignor: CELLTECH LIMITED
  • Assignee: ALUSUISSE HOLDINGS A.G.
  • Correspondent: not retrieved. (Patent-face prosecution agent of record is SPENCER, FRANK & SCHNEIDER, per FreePatentsOnline — but a patent-face agent is not the assignment correspondent, and I have no evidence the two are the same. Flagged as a lead only, not a finding.)
  • Context: Acquisition / strategic divestiture — Celltech's sale of its Biologics (expression-technology) business to the Alusuisse group; a real arm's-length business sale, not a shell transfer.

• Executed date not retrieved / recorded 2001-07-23 — Reel not retrieved / not retrieved

  • Conveyance: Merger
  • Assignor: ALUSUISSE HOLDINGS, AG
  • Assignee: ALUSUISSE LONZA GROUP, AG
  • Correspondent: not retrieved.
  • Context: Internal corporate reorganisation — upstream holding-company merger consolidating the Alusuisse-Lonza group.

• Executed date not retrieved / recorded 2001-07-23 — Reel not retrieved / not retrieved

  • Conveyance: Assignment of assignors' interest
  • Assignor: ALUSUISSE LONZA GROUP, AG
  • Assignee: LONZA GROUP, AG
  • Correspondent: not retrieved.
  • Context: Spin-off / internal reorganisation — placement of the chemicals-and-life-sciences assets (including Lonza Biologics) into the demerged Lonza Group Ltd/AG, which listed independently on the Swiss exchange in 1999–2001. This is the link that terminates the chain at the operating entity that later asserted the patent.

• Executed date not retrieved / recorded 2006-08-01 — Reel not retrieved / not retrieved

  • Conveyance: Security Agreement
  • Assignor: PHARMATHENE, INC.
  • Assignee: MPM BIOVENTURES III-QP, L.P., AS ADMINISTRATIVE AGENT
  • Correspondent: not retrieved.
  • Context: ⚠️ ANOMALY — appears to be a mismapped or errant record. PharmAthene, Inc. is a US biodefense company with no discernible relationship to Lonza, Celltech, or the hCMV-MIE/GS patent family, and MPM BioVentures was a PharmAthene investor — this record reads like a venture-lender security interest over PharmAthene's own patent portfolio that has been mis-associated with US 5,658,759 in the aggregated data. I cannot confirm from the available data whether the USPTO actually recorded a lien against this patent or whether this is an indexing artifact. Do not rely on it as an encumbrance on this patent. It is flagged here because it will contaminate any automated ownership scrape.

No post-2001 transfer away from the Lonza group was found. In particular, the separately listed current assignee "Rio Tinto Switzerland AG" is, in my assessment, very likely a data artifact of the Alusuisse → Alcan (2000) → Rio Tinto Alcan (2007) corporate chain, since the biologics assets had already been demerged into Lonza Group by 2001. I found no recorded assignment to Rio Tinto and I am not asserting one. Treat the Rio Tinto listing as unverified.

Timeline diagram

timeline
    title Ownership and assertion of US 5658759
    1987 : GB priority application filed
    1988 : PCT application filed
    1989 : US national stage filed
    1996 : Divisional application filed
         : Celltech Biologics sold to Alusuisse
    1997 : Patent issued 19 Aug
         : Recorded assignment to Alusuisse Holdings AG
    2001 : Merger creates Alusuisse Lonza Group AG
         : Recorded assignment to Lonza Group AG
    2006 : Security agreement recorded involving PharmAthene
    2007 : Lonza sues Northwest Biotherapeutics
    2009 : Patent expires 28 Apr

NPE / troll-pattern signals

1. Shell-entity transfer — NOT PRESENT. All four recorded links run between named operating corporations (Celltech Limited → Alusuisse Holdings A.G. → Alusuisse Lonza Group AG → Lonza Group AG). No "IP/Holdings/Ventures" licensing vehicle appears, and the chain terminates at a company that actually operates the technology. The sole outlier name in the data (the 2006-08-01 PharmAthene/MPM BioVentures security agreement) is an apparent data-mapping anomaly, not a shell transfer of this patent.

2. Known asserter in the chain — NOT PRESENT. No current or prior assignee matches a published NPE/asserter list (Acacia, Marathon, IV, IPNav, Wi-LAN, Conversant/Mosaid, Vringo, Pendrell, Round Rock, MPHJ, Lumen View, Spangenberg entities, etc.). Lonza Group AG / Lonza Biologics is a Swiss-listed operating life-sciences company.

3. Repeat correspondent across the chain — UNCLEAR (not assessable). I could retrieve no correspondent data at all. This signal cannot be scored without Assignment Center access. Note for follow-up: the patent-face prosecution firm is Spencer, Frank & Schneider; whether that firm also acted as recording correspondent on any link is unknown. No finding.

4. Cascading transfers — NOT PRESENT. Four recorded events over roughly ten years, each with a distinct documented corporate rationale (business-unit sale; holding-company merger; demerger). Nothing resembling chained LLC-to-LLC transfers within 24 months.

5. Pre-litigation transfer — NOT PRESENT. The 2007 assertion was brought by Lonza Group AG, which had held the patent by recorded assignment since 2001-07-23 — approximately six years before suit. There is no assignment in the window shortly before the first suit.

6. Bankruptcy fire-sale — NOT PRESENT. The 1996/1997 transfer was a strategic divestiture of an operating division by a then-solvent Celltech; Alusuisse was a solvent acquirer. Celltech's later absorption into UCB (2004) occurred after the patents had already left the group.

7. Privateering — UNCLEAR / partially indicated. Lonza is not a classic non-practising entity, but its posture is worth noting: Lonza's biologics arm is principally a contract manufacturer and technology licensor, so its 2007 suit against Northwest Biotherapeutics over the DCVax products reads as an operating company enforcing a licensed platform technology against a user, rather than a pure troll shakedown. This is an operating-company enforcement action; I flag only that a service/licensing business enforcing a platform patent sits near the privateering boundary.

8. Defensive aggregator — NOT PRESENT. The chain does not terminate at RPX, AST, LOT, Unified Patents, or OIN. The patent lapsed by terminal expiry (see below), not by defensive acquisition.

Terminal event: the patent's anticipated expiration is 2009-04-28 — exactly 20 years from the 1989-04-28 US national-stage filing of parent 07/339,615 — and its status is Expired – Fee Related. Any assertion value is therefore historical only.

Verdict

Operating-company assertion. The recorded chain runs Celltech Limited → Alusuisse Holdings A.G. (recorded 1997-11-10) → Alusuisse Lonza Group AG (merger, recorded 2001-07-23) → Lonza Group AG (recorded 2001-07-23), and terminates at a genuinely operating life-sciences company that commercialises the claimed CHO/GS cell technology through Lonza Biologics; that same entity asserted US 5,658,759 in the 2007 Lonza v. Northwest Biotherapeutics litigation, roughly six years after acquiring it — the classic posture of a practising company suing an unlicensed user, not a shell-to-shell NPE chain. No shell-entity transfer, no known-asserter in the chain, no pre-litigation transfer, and no bankruptcy sale are present. The one apparent abstention is signal #3 (correspondent recurrence), which is [unclear due to data inaccessibility], not negative — if you need it resolved, run the patent number directly in the USPTO Assignment Center and read the correspondent fields, and treat the 2006-08-01 PharmAthene/MPM BioVentures entry as needing manual verification before it is relied upon.

Verification link: https://assignmentcenter.uspto.gov/ (search patent number 5658759); mirror: https://assignment.uspto.gov/patent/index.html.

I'll research the prior art cited against US 5,658,759, starting with the patent's own citation records.

Let me retrieve the full citation list from the patent records.

Let me pull the complete reference lists (patent + non-patent) from the patent's front page.

I hit my tool-call limit before I could retrieve every citation table, so I'll report what I verified, flag what I could not verify, and give the §102 analysis on the references I did confirm.

Prior art analysis — US 5,658,759 ("Recombinant DNA expression vectors")

0. Verification status and honest caveats

• I confirmed the patent number 5658759 is the Bebbington/Celltech "Recombinant DNA expression vectors" patent via Google Patents US5658759A, Justia 5658759, and FreePatentsOnline 5658759. No other patent shares that number.
• What I could verify: the front-page U.S. Patent Documents list (a single entry) and the non-patent literature list, plus the references discussed in the specification body.
• What I could NOT verify: an explicit front-page Foreign Patent Documents table for this patent. The search-result snippets for both Justia and FPO did not surface one, and I could not complete a page fetch. I therefore treat the foreign references below (EP 260148, WO 87/04462) as applicant-discussed/specification-cited art, not as confirmed examiner citations, and I label them as such.
• I am not asserting a definitive legal conclusion. Anticipation under §102 requires a single reference disclosing every element of a claim; my per-claim notes are analytical, not adjudicated.

1. The patent's own claim set (for the anticipation matrix)

From the issued claims (quoted in the full text and confirmed on FPO/Justia):

The critical novelty hook for claim 1 is the phrase "complete 5'-untranslated region including the first intron" of hCMV-MIE. Any §102 attack must show a single reference that already put that full element (promoter + enhancer + complete 5'-UTR + first intron) into a transfected host cell operably linked to a heterologous coding sequence.

2. Patent citations

2.1 U.S. Patent Documents (the only one on the front page)

Note: the Justia "Referenced Cited" table for this patent shows only this one U.S. patent document. I found no additional U.S. patent citations in the front-page record I could retrieve.

2.2 Foreign/specification-cited references (candidate prior art; citation status unconfirmed)

These are cited in the body of US 5,658,759 as the state of the art, which is why they are the natural prior-art set to test.

(a) EP 260148 — "Published European Patent Application No. 260148," directed to a method for the continuous production of a heterologous protein.

• Description (per the 5658759 specification): the expression vectors there contain part of the 5'-untranslated region of the hCMV-MIE gene as a stabilising sequence, placed in the 5'-UTR of the heterologous gene; the specification says its teaching is "again that the natural 5'-untranslated region of the gene is essential for translation."
• §102 relevance: This is the closest commercial-competitor art, but its own disclosure is partial 5'-UTR used as a stabiliser, with translation driven off the heterologous gene's natural 5'-UTR — i.e., it does not disclose the "complete 5'-UTR including the first intron" element of claim 1. On the face of the specification's characterization, it does not anticipate claims 1–10; it is more naturally §103 art. Date caveat: EP 260148 published in early 1988, i.e., after the 1987-07-23 priority date, so it is likely not §102 prior art at all against these claims (and, being a foreign publication, it is not §102(e) art). I could not independently confirm its exact publication date within my search budget.

(b) WO 87/04462 — Celltech, cited in the specification as the source of pSV2.GS ("The production of pSV2.GS is described in published International Patent Application No. WO 8704462").

• Description: glutamine-synthetase (GS) vector/selection system used to build pCMGS.
• §102 relevance: Relevant to the GS selectable-marker aspects and to the makeup of the claim-6 plasmids (pCMGS). It would not, by itself, disclose the hCMV-MIE complete-5'-UTR element. Potential §102/§103 art only in combination with an hCMV-MIE reference; standalone anticipation of claims 1–10 is not supported on the disclosure characterized in the patent.

3. Non-patent literature (examiner-cited front page)

These are the substantive §102/§103 references. All are pre-priority except where noted.

Source for items 1–21: the "[U.S. Patent Documents / References Cited]" and "Non-patent literature cited in the description" tables on Justia and FreePatentsOnline.

4. Bottom line on anticipation

On the record I could verify, no single reference anticipates any of claims 1–10:

• Claim 1 requires the complete hCMV-MIE 5'-UTR including the first intron, operably linked to a heterologous coding sequence, inside a transfected host cell. The nearest references (Spaete & Mocarski 1985; EP 260148; Foecking & Hofstetter 1986) disclose only the promoter/enhancer and, at most, part of the 5'-UTR — and Spaete & Mocarski and EP 260148 expressly rely on the heterologous gene's own 5'-UTR for translation. That is the gap the 5658759 claims exploit.
• Claim 5 (hCMV-MIE translational initiation signal) and claim 6 (pCMGS / pEE6.hCMV / pHT.1) are narrower still, and no cited reference names these constructs.
• Claims 2 and 7–10 are process claims that ride on the vector/host-cell limitations above; they fail or stand with claim 1.
• Claims 3, 4 add a restriction site and a direct promoter→coding-sequence fusion; the specification itself frames the direct fusion as the discovery, and none of the cited art is described as fusing the complete hCMV-MIE 5'-UTR (first intron included) directly to a coding sequence.

The realistic role of the cited art is §103 obviousness (hCMV promoter/enhancer + a mammalian expression vector + an amplifiable marker), not anticipation. The single U.S. patent citation (US 4,769,326, Rutter) and the EP 260148 / WO 87/04462 foreign references are the items most worth pulling full-text for before making any anticipation call — and, per my caveats above, US 4,769,326 issued after the priority date and EP 260148/none of the foreign publications can serve as §102(e) art, so their real bite would be under §103.

Open items I could not close within my tool budget: (i) the patent's front-page Foreign Patent Documents table, if one exists; (ii) the actual disclosure of US 4,769,326; (iii) the precise publication date of EP 260148. If you want, I can target those three specifically next.

I'll verify the key prior-art references and their dates before building the §103 combinations.

Two verification gaps remain: the sole U.S. patent citation (US 4,769,326, "Rutter") and whether any pre-July-1987 art already used the hCMV-MIE intron in an expression cassette.

Obviousness analysis — US 5,658,759 under 35 U.S.C. § 103

Two housekeeping flags before the analysis.

1. Date conflict. This prompt states the current date as April 26, 2026. The authoritative full text supplied for this patent was fetched 2026-09-27T19:21:24Z and the earlier sections in this file were written to that date. I analyze to the 1987 priority date (which is what § 103 turns on) and note the conflict rather than silently harmonizing it.
2. A verified contradiction with the Prior Art section. The Prior Art section treated EP 260148 as "the closest commercial-competitor art" and flagged its publication date as unconfirmed. I have now confirmed the date, and it changes the analysis materially: EP 260148 A2 published in March 1988 (A3 published 1989-06-07; priority 1986-09-12; filing 1987-09-11; Genentech/Gorman, the EP counterpart family of US 5,024,939). That is after the 1987-07-23 priority date, so EP 260148 is not § 102(a) art, and it cannot be § 102(b) art either (see the § 102(b) date discussion below). Its contents are still useful to us because it marshals pre-1987 intron/splicing literature — but the reference itself cannot be combined into a § 103 rejection. Source: EP0260148A2 full text, PubChem EP-0260148-A3 record.

I am not rendering a legal opinion or a validity conclusion; this is a technical-analyst assessment.

1. The legal frame and the level of ordinary skill

Governing law. Priority is 1987-07-23, so this is a pre-AIA § 103(a) patent. The Graham v. John Deere factors govern, and KSR Int'l Co. v. Teleflex Inc., 550 U.S. 398 (2007) applies its "predictable use of prior art elements according to their established functions" and "finite number of identified, predictable solutions" reasoning.

PHOSITA (proposed). A molecular biologist holding a Ph.D. (or M.S. plus ~3–5 years) with practical experience constructing mammalian expression vectors, transfecting CHO/CO1 cells by calcium phosphate or DEAE-dextran, and using a dominant selectable/amplifiable marker (DHFR/MTX or GS/MSX). This is the skill level reflected by the cited literature itself — the examiner-cited references are routine vector-construction papers, not exotica.

Critical date. The GB priority application was filed 1987-07-23; PCT/GB88/00602 was filed 1988-07-22 (exactly at the 12-month Paris anniversary); the US national stage 07/339,615 was filed 1989-04-28. Two consequences:

• § 102(a): references published before 1987-07-23 are in. Everything I rely on below is in.
• § 102(b): under 35 U.S.C. § 363 the international filing date (1988-07-22) is the effective US filing date, making the § 102(b) critical date 1987-07-22. So WO 87/04462 (published 1987-07-30) and EP 260148 (March 1988) both fall inside the grace period and are not statutory-bar art. WO 87/04462 is additionally not § 102(e) art under the law in force for 1987–89 filings (pre-AIPA § 102(e) covered only US patents), and it is the applicant's own Celltech work in any event.

This is favorable to the patentee and should be stated plainly rather than glossed.

2. What each claim actually needs, and what the art supplies

Claim 1 is the only independent claim and is the whole ballgame; claims 3–6 are its dependent species, and claims 2 and 7–10 ride entirely on it.

Stenberg 1984 is the pivot of the whole analysis and it was already on the face of the patent. Source: J. Virol. 49:190-199, PubMed 6317889.

3. The applicant's own admission does most of the work

The specification (Example 1) states, in the applicant's own words:

"The Pst-I fragment of hCMV … contains the promoter-enhancer and most of the 5'-untranslated leader of the MIE gene including the first intron. The remainder of the 5'-untranslated sequence can be recreated by attaching a small additional sequence of approximately 20 base pairs."

Read against Spaete & Mocarski (who used that very PstI fragment as a promoter for heterologous genes), this is a specification admission that (a) the prior-art promoter fragment already contained the first intron, and (b) the only missing piece was ~20 bp plus a restriction-site junction. Admissions in the specification are usable as prior art against the applicant's own claims. This does more damage to claim 1 than any single reference, because it collapses the asserted novelty hook — the "complete … including the first intron" element — into roughly two decades of nucleotides and a synthetic linker.

The specification makes two further admissions that are fatal to the narrow claims:

• "Many eukaryotic genes contain an NcoI restriction site (5'-CCATGG-3') overlapping the translation start site, since this sequence frequently forms part of a preferred translation initiation signal 5'-ACCATGPu-3'." → the claim-5 NcoI variant is conceded conventional (Kozak, Cell 41:283-292 (1986)).
• pCMGS is described as the synthetic oligomer serving "merely as a convenient PstI–NcoI adaptor" → the claim-3 "restriction site to facilitate insertion" is conceded to be a convenience feature.

4. Combination 1 — the primary § 103 case against claim 1

Combination: Boshart Cell 41:521-530 (1985) + Spaete & Mocarski J. Virol. 56(1):135-143 (1985) + Stenberg J. Virol. 49(1):190-199 (1984), optionally with Foecking & Hofstetter Gene 45:101-105 (1986) and Pasleau Gene 38:227-232 (1985) as secondary teaching, and Kaufman J. Mol. Biol. 159:601-621 (1982) for the CHO host.

Why a PHOSITA would have combined them:

1. Known need / stated problem. The field's aim was maximal polypeptide yield (the patent says so itself). Boshart and Foecking & Hofstetter independently established the hCMV-MIE enhancer-promoter as the strongest available mammalian element; Foecking & Hofstetter's title — "Powerful and versatile enhancer-promoter unit for mammalian expression vectors" — is itself the design incentive. There is a direct, articulated reason to put it in an expression vector.
2. The lead reference already did 90% of the work. Spaete & Mocarski placed the same PstI fragment upstream of heterologous coding sequences and got expression. The only unfilled element is completing the leader.
3. The complete leader was a known, sequenced, finite article. Stenberg 1984 disclosed the exon/intron structure and the splice junctions of the MIE leader. Reconstructing the complete 5'-UTR therefore required no discovery — only retrieval of published sequence plus a synthetic oligomer. KSR: where the elements are known and the result is the predictable use of each for its established function, the combination is obvious.
4. A known mechanism supplied the reason to retain the intron. Mulligan Nature 277:108-114 (1979) (recombinant genomes with 5' leader splice sites produce hybrid mRNA and "substantial quantities" of protein; splicing-defective mutants fail to produce discrete mRNAs), Hamer & Leder Cell 1979 ("splicing is a prerequisite for stable RNA formation" — as quoted in the EP 260148 discussion), and Gruss PNAS 76:4317-4321 (1979) (intervening-sequence-minus mutant makes no detectable protein) collectively taught that an intact, spliceable 5' leader is important for mRNA accumulation. Retaining the natural first intron rather than deleting it is the predictable application of that teaching. All three were already cited on the 5,658,759 face.
5. Reinforcing teaching that introns can carry regulatory elements: Gillies Cell 33:717 (1983), Banerji Cell 33:729 (1983) and Queen & Baltimore Cell 33:741 (1983) localized enhancer activity inside an intron; Fisher JBC 260:1122 (1985) reported an intron in the tPA mRNA leader before the coding sequence.
6. The "finite predictable solutions" point. Given a known promoter/enhancer and a known leader, the artisan's choice is essentially binary: truncate the leader (as Spaete & Mocarski did, incidentally) or keep it whole. That is not an unbounded field of candidates; it is the classic KSR "obvious to try" posture.

Combination 1A (fallback, avoiding any date argument on Boshart). Boshart Cell 1985 + Stenberg 1984 + Thomsen 1984 + Foecking & Hofstetter 1986 + Kaufman 1982. Spaete & Mocarski is not strictly necessary, since the applicant's own admission supplies the "prior fragment contained the first intron" fact.

5. Combination 2 — claim 4 (direct linkage) and claim 5 (native initiation signal)

Combination: Combination 1 + Kozak Cell 41:283-292 (1986) and PNAS 83:2850-2854 (1986) + Pelletier & Sonenberg Cell 40:515-526 (1985).

Motivation. The specification itself concedes the art knew that using a hybrid 5'-UTR was unreliable: "the presence of particular 5'-untranslated sequences can lead to poor initiation of translation (Kozak … and Pelletier and Sonenberg …)." A PHOSITA seeking reproducible, gene-independent expression therefore had an affirmative reason to delete the variable heterologous leader and drive translation from a single, characterized viral leader at a defined initiation codon — i.e., exactly the direct fusion of claim 4. Kozak supplied the consensus (ACCATGPu) and thus the reason to site the ATG in that context; the specification concedes that an NcoI site at the ATG is conventional. Claim 5 (retaining the hCMV-MIE translational initiation signal) is inherent in using the complete natural leader, and the G→C alteration at −1 is an admitted routine optimisation.

6. Combination 3 — claim 3 (restriction site)

Motivation: self-evident. Every expression vector needs a cloning site; the patent concedes the synthetic oligomer was used "merely as a convenient PstI–NcoI adaptor." No separate reference is required; this is the "arrangement of parts" level of ordinary creativity.

7. Combination 4 — claim 6 (pCMGS, pEE6.hCMV, pHT.1)

This is the weakest limb of the challenger's case, and I would grade it "plausible but contested."

Why the combination is arguable. Each component is a known element used for its known purpose (backbone/selection/promoter/cargo), and KSR treats "combining familiar elements according to known methods yielding predictable results" as obvious. Notably, the specification attributes no unexpected property to the specific plasmid architecture — the asserted benefit (high-level expression) is attributed to the leader, which is claim 1's subject matter. That is fatal to any attempt to build a separate nonobviousness argument on claim 6.

Why it is nonetheless contestable. A claim to a specific plasmid is a claim to a specific molecule; where the specification shows no comparative data against an obvious alternative construct, a challenger must still articulate why these junctions and this backbone were the predictable outcome. Expect this claim to survive or fall with claim 1 rather than on its own merits.

8. Claims 2 and 7–10 (production processes)

Combination: any of Combinations 1–4 + the general heterologous-expression literature already of record — Pasleau 1985, Gray Nature 295:503-508 (1982), Gething Nature 293:620-625 (1981), Leonard Nature 311:626-631 (1984), McLean PNAS 83:2335-2339 (1986) — each of which teaches culturing a transfected eukaryotic host and recovering the secreted protein; and US 4,769,326 (UC Regents, "Expression linkers," issued 1988-09-06), which claims an "expression system capable of effecting the expression of the recombinant DNA sequence … in a suitable host cell" and a "method for producing a recombinant protein … compris[ing] culturing the cells." US 4,769,326 is unavailable under § 102(a)/(b) (it issued after the priority date) but is a § 102(e) candidate if its application filing date precedes 1987-07-23 — which is plausible for a 1988 grant and should be pulled from the assignment/PALM record. Either way it is only background: nothing in it touches hCMV.

The step of "culturing the cell" adds nothing patentable to claim 1 — it is the universal, disclosed utility of every expression vector in the art. Claims 2 and 7–10 stand or fall with the claims they depend from.

9. The patentee's counterarguments, and how they fare

(a) Teaching away. The patent's core narrative is that the art "taught that the natural 5'-untranslated region of the gene is essential for translation" (referring to Spaete & Mocarski and to EP 260148). This is the strongest nonobviousness argument available and it is a real one — but it is weaker than it looks:

• Using a different design is not teaching away. A reference must criticise, discredit or discourage the claimed approach. Spaete & Mocarski simply made a different construct; it did not disparage the complete-leader approach.
• The "teaching away" case depends heavily on EP 260148 — which is not prior art. A teaching-away argument must be grounded in § 102 art, so the patentee cannot lean on EP 260148's "natural 5'-UTR is essential" language.
• Worse for the patentee, the art was genuinely split, and some of it cut against introns: Reddy J. Cell. Biochem. Suppl. 10D:154 (1986) found that including introns reduced expression and concluded introns "were not an essential part of vectors," and Hall et al. reportedly made the same observation. That is genuine teaching-away material against claim 1's first intron (and a challenger must confront it — e.g., by arguing those constructs used heterologous/artificial introns in a different positional context rather than the natural, spliced leader of the promoter actually being used).

(b) Unexpected results. The patent's data are genuinely striking: 3×10⁸ molecules TIMP/cell/day at ~a single integrated copy, rising to 3×10⁹ after one MSX amplification round, plus consistent PA-1, PA-2 and light-chain expression. Rebuttals a challenger would press:

• The data measure absolute yield and copy-number efficiency, not superiority over the closest prior-art construct (a Spaete & Mocarski-style vector with the heterologous leader). Without a head-to-head comparison, the "unexpected" character is under-demonstrated, and commensurateness-in-scope attacks apply.
• The asserted effect — strong, promoter-driven expression — is precisely the known function of the hCMV MIE enhancer/promoter. Under KSR, a result flowing from an element's established function is not automatically nonobvious.
• The specification's emphasis that "it is almost impossible to predict whether any particular vector or cell line or combination thereof will lead to a useful level of production" is a broad unpredictability argument. It is rhetorically useful but cuts both ways, because it is equally consistent with the routine, empirical vector-optimisation work that the Mulligan/Hamer/Gruss/Reddy literature instructed.

(c) Objective evidence of commercial success / industry adoption. The hCMV-MIE-leader/GS platform became an industry standard, but the nexus problem is severe: the commercial success is largely attributable to the GS/MSX selection system and CHO host platform, neither of which is an element of claim 1. The applicant's own WO 87/04462 already described the GS vector/selection system, so the platform's value cannot be credited to the claimed single feature (completeness of the leader).

10. Overall assessment by claim

One claim-construction caveat that matters enormously. Claim 1 recites the "complete 5' untranslated region," while the abstract and summary recite a "substantially complete" one. If "complete" is construed strictly, the case is § 103 (no single reference reaches it). If it is construed to cover a "substantially complete" leader, then Spaete & Mocarski's PstI fragment — which the applicant admits contained "most of" the leader including the first intron — moves the analysis toward § 102 anticipation. That is a validity knife-edge the patentee created for itself in the specification, and it is where I would focus if this patent ever had to be defended.

11. Confidence, gaps, and what I could not verify

• Verified this round: Stenberg 1984's disclosure of the MIE leader exon/intron structure and splice junctions (J. Virol.); EP 260148's priority/filing/publication dates and its intron/splicing discussion (Google Patents); US 4,769,326's title ("Expression linkers"), assignee (The Regents of the University of California) and claim 6/7 subject matter (FreePatentsOnline).
• Open items (unchanged from the Prior Art section, and they matter): the front-page Foreign Patent Documents table for 5,658,759 was never retrieved; US 4,769,326's application filing date is needed before it can be called § 102(e) art; Boshart Cell 41:521-530's exact disclosure of the leader region (vs. the enhancer only) should be read in full before finalising Combination 1.
• Not verified: the precise filing date and language of WO 87/04462 (which I treat as non-art regardless), and whether any 1987 prosecution statement addresses the "complete" vs. "substantially complete" gap in claim 1.
• Practice note: as the earlier sections establish, this patent is expired (anticipated expiration 2009-04-28) with no PTAB record. This § 103 analysis therefore matters for validity opinions, licensing diligence, invalidity-contentions in a family dispute, and double-patenting analysis against the sibling US 5,591,639 — not for defending live infringement exposure, which § 286 forecloses.