Introduction
CAR-T therapies shaped how the industry thinks about patenting engineered cells, but the field has moved well beyond them. Today’s pipelines include TCR-engineered T cells, tumor-infiltrating lymphocytes (TILs), NK and CAR-NK cells, regulatory T cells (Tregs), gamma-delta T cells, engineered macrophages, iPSC-derived products, mesenchymal stromal cells and gene-edited hematopoietic stem cells. These platforms differ biologically, but they share a legal feature: most of what makes them inventive is written in nucleic acid and amino acid sequences, such as receptors, guide RNAs, editing templates, cytokine constructs, safety switches and vector elements.
The sequence listing is where that information must be formally disclosed. Errors are common and they are costly. A missing or noncompliant listing can trigger notices and deadlines and a listing that omits sequences you later want to claim cannot be fixed without adding new matter.
This article covers the sequence listing framework, which sequences non-CAR-T cell therapy applications must include, tricky representation issues, filing mechanics and how the listing interacts with written description, enablement and eligibility. It focuses on U.S. practice under 35 U.S.C. § 112, 37 C.F.R. §§ 1.831–1.839 and the MPEP, with a comparative section on the EPO, PCT and India.
Part 1: Why the Sequence Listing Matters
It is part of the disclosure
A patent must describe and enable the invention (35 U.S.C. § 112(a)). For sequence-defined subject matter, the sequence is the structure. When a claim recites “a TCR comprising the CDR3 of SEQ ID NO: 5,” the sequence listing and the description together are what support that claim.
Consequences of getting it wrong
- Formal deficiencies. The USPTO issues a notice requiring a compliant listing, with a time period for response. Failure to respond can lead to abandonment.
- No new matter. A sequence not disclosed at filing cannot be added later. Continuations and divisionals can only claim what the original application supports.
- Weak claim support. Vague sequence disclosure invites written description and enablement attacks (see Part 7).
- Cost and delay. Late corrections often mean additional filings, statements and, in some offices, fees.
Part 2: The Governing Standard: WIPO ST.26
The transition from ST.25
WIPO Standard ST.26 replaced ST.25 for applications filed (or, for PCT, with an international filing date) on or after July 1, 2022. ST.26 is an XML-based standard. Applications filed before that date continue under the older ST.25 rules (37 C.F.R. §§ 1.821–1.825), but any later-filed application, including in many cases a continuation, divisional, or national-stage entry, generally must comply with ST.26.
WIPO has revised ST.26 since its adoption, so applicants should always use the current version of the standard and of the WIPO Sequence tool.
Key differences that matter for cell therapy
| Feature | ST.25 (legacy) | ST.26 |
| File format | Plain text (.txt) | XML (.xml) |
| Features and qualifiers | Loosely controlled | Controlled vocabulary with mandatory qualifiers for certain features |
| Free text | Numeric-tagged fields such as <223> | Confined to defined qualifiers, such as “note” |
| Scope | Natural residues; limited handling of modifications | Also covers D-amino acids and nucleotide analogues |
| Creation and validation | Various tools | WIPO Sequence tool and Office-side validation |
What triggers the listing requirement
In general, the listing must include:
- Nucleotide sequences of ten or more specifically defined nucleotides.
- Amino acid sequences of four or more specifically defined amino acids.
“Specifically defined” means residues that are not ambiguous placeholders such as “n” or “X.” Branched sequences and certain other structures are outside the standard. The definitions are technical, so check the standard’s text rather than relying on rules of thumb.
The consequence for cell therapy is that many short sequences need identifiers. A five-residue linker, a nine-residue epitope tag, or a twenty-nucleotide guide RNA spacer all meet the thresholds.
Part 3: Sequence-Bearing Elements by Modality
CAR patents make sequence obligations easy to spot because the receptor is a fusion protein. Other modalities tend to bury sequences in editing reagents, assays and vectors. The following table shows commonly overlooked elements. It is illustrative, not exhaustive.
| Modality | Sequence-bearing elements that often need SEQ ID NOs |
| TCR-engineered T cells | TCR α and β variable and constant regions; CDR1–3; engineered constant domains (for example, added cysteines or murinized regions); codon-optimized cDNAs; target peptide epitopes (often 8–15 residues); HLA-restricting allele references; endogenous TRAC/TRBC knockout guide sequences |
| TIL therapies | Tumor-reactive TCR clonotypes; engineered cytokine constructs (for example, membrane-bound IL-15); PD-1 or other knockout guides; TCR-sequencing primers if recited |
| NK, CAR-NK and iPSC-derived NK | Engineered CD16 variants; NKG2D chimeras; IL-15/IL-15Rα fusions; knockout guides (for example, for CISH); HLA-E single-chain constructs; safety switches |
| Regulatory T cells | FOXP3 constructs; CAR or TCR elements; enhancer and promoter sequences; assay primers (for example, for demethylation assays) |
| Gamma-delta T cells | Vγ/Vδ receptor chains; engineered receptors; cytokine constructs |
| Macrophages and myeloid cells | Chimeric receptors; cytokine or phagocytosis-modulating constructs |
| iPSC, ESC and MSC products | Reprogramming factor constructs and mRNAs; guide RNAs; knock-in donor templates; hypoimmunogenic edit cassettes (for example, B2M/CIITA targeting, CD47 overexpression); lineage-reporter promoters; suicide genes |
| Gene-edited HSCs | Guide RNAs targeting enhancers or coding regions; base or prime editor guide sequences; homology arms; vector components |
| Dendritic cell and antigen-loaded products | Neoantigen peptides; encoding mRNA sequences |
Elements common to nearly all engineered cells
These often appear in the description and are easy to miss:
- Signal peptides, hinge and transmembrane domains, costimulatory and signaling domains
- Linkers (for example, glycine-serine linkers), which are often short but exceed four residues
- Self-cleaving 2A peptides (P2A, T2A and similar)
- Epitope tags and marker or safety-switch sequences (for example, truncated surface markers, inducible caspase constructs, or tag-and-epitope hybrid safety markers)
- Promoters, enhancers, LTRs, polyA signals, WPRE elements and transposon inverted repeats
- Primers, probes, barcodes and sequencing adaptors used in assays
If a sequence meets the length threshold and appears in the description, claims, or drawings, it needs a SEQ ID NO, whether or not it is inventive.
Part 4: Gene Editing Components and Vectors
Editing is central to allogeneic, “off-the-shelf,” and hypoimmunogenic therapies and it produces many short sequences.
Guide RNAs and target sites. Both the guide sequence (typically about 20 nucleotides) and the genomic target site meet the ten-nucleotide threshold, so each should be listed. Where the guide and target are related, keep the relationship clear through consistent annotation.
Base editors, prime editors and nucleases. The editor proteins are usually well known, but engineered variants (mutations, fusions, linkers) may need their own sequences. Prime editing guide RNAs (pegRNAs) include primer-binding and template regions that should be described and listed accurately.
Donor templates and homology arms. Knock-in cassettes, including safety-switch or receptor insertions at safe-harbor loci, are long sequences that must be listed, with features marking arms, promoters and coding regions.
Vectors. Lentiviral, gamma-retroviral, AAV and transposon-based constructs (including Sleeping Beauty and piggyBac systems) may be disclosed as full sequences. Annotating features correctly is essential in an XML listing (see Part 5).
Chemically modified guides. Guide RNAs often carry 2′-O-methyl, phosphorothioate, or other modifications. ST.26 has specific rules for representing modified nucleotides and analogues and modifications must be handled through the standard’s feature mechanism rather than ad hoc notes. Confirm treatment against the standard before assuming a modified guide is or is not covered.
RNA sequences. RNA is represented under specific conventions in ST.26, including how uracil is symbolized and how the molecule type is annotated. Follow the standard’s annexes rather than the conventions used in the description.
Part 5: Tricky Representation Issues
Fusion proteins and multi-domain constructs
List the full fusion as a SEQ ID NO and annotate its parts with features such as signal, domain and region. If a claim also recites an individual domain (for example, a specific costimulatory domain), that domain should be listed as its own sequence when it meets the length threshold.
Coding sequences and encoded proteins
If you annotate a coding sequence, keep nucleotide and amino acid listings internally consistent. A polypeptide encoded by a listed CDS generally should also be listed as an amino acid sequence and codon-optimized variants that differ in sequence are different SEQ ID NOs.
Variable and generic sequences
Formulas with variable residues (for example, “X1-X2-X3”) are handled using placeholder symbols plus features that describe the variation. A sequence made only of unspecified residues, or with too few specifically defined residues, falls outside the requirement. Use the feature mechanism to describe permitted variants and consider whether you also want individual species listed to support fallback claims.
Sequences in figures, tables and text
Every disclosed sequence that meets the threshold must be identified by SEQ ID NO wherever it appears: description, claims and drawings. Plasmid maps, alignment figures and sequence tables are frequent offenders. Where a figure shows a sequence, include its identifier in the figure or its legend.
Primers, probes, barcodes and libraries
Experimental sections often recite primers, probes, unique molecular identifiers and barcodes. These count if they meet the length threshold and are recited in the application. Very large libraries (for example, pooled guide libraries) raise practical questions about how to present them; consult current USPTO guidance and consider whether the library needs to be recited at all.
Natural sequences and database references
Relying on a gene name or database accession number in place of a sequence is risky. Database entries change over time and MPEP § 608.01(p) restricts incorporation by reference of essential material. For anything material to the invention, disclose the sequence itself in the listing, even if it is a known human sequence.
Feature keys and qualifiers
Choose feature keys from the standard’s controlled vocabulary (for example, keys such as CDS, promoter and misc_feature for nucleotides and keys such as SIGNAL, DOMAIN, REGION, VARIANT and MOD_RES for amino acids) and supply mandatory qualifiers, such as the organism and molecule type on source features. Missing mandatory qualifiers are among the most common validation failures.
Part 6: Filing Mechanics
United States
- Format and submission. File the listing as an ST.26-compliant XML file through Patent Center, as part of the application rather than as a PDF or image.
- Incorporation by reference. The specification typically must include a statement incorporating the XML file by reference, identifying its file name, creation date and size in bytes.
- Page-count treatment. An XML listing is generally excluded from the application size fee calculation.
- Validation. Validate the file with the WIPO Sequence tool before filing. The USPTO also runs validation on submission and errors may produce a notice.
- Late-filed listings. A listing submitted after the filing date must be accompanied by a statement that it contains no new matter and adds nothing beyond the original disclosure.
- Priority applications and continuing applications. If a priority document or parent application used ST.25, the later-filed application generally must include an ST.26 listing. Converting the older file using WIPO tools can introduce differences, so review the converted output carefully and include the required no-new-matter statement.
PCT
Under PCT Rules 5.2 and 13ter, a sequence listing forms part of the description when filed with the international application. A listing furnished later is not part of the disclosure and is considered only for search purposes and a late-furnishing fee may apply. Since PCT is the usual route for global cell therapy filings, the sequence listing should be complete on day one.
Quality control before filing
- Cross-check every SEQ ID NO in the description, claims and figures against the XML file.
- Confirm that every sequence meeting the length threshold in the text has an identifier.
- Validate the XML and resolve all errors and warnings.
- Confirm that the same SEQ ID NOs are used consistently across family members where possible.
Part 7: Substantive Law: Where Sequences Meet Patentability
Written description and enablement
Cell therapy claims are frequently written in functional terms, such as “a T cell comprising a receptor that binds antigen X.” The Federal Circuit has shown little patience for that approach when the specification does not adequately disclose the claimed range of structures.
- In Juno Therapeutics, Inc. v. Kite Pharma, Inc., 10 F.4th 1330 (Fed. Cir. 2021), the court upheld invalidity of claims to CAR-T cells with a binding element specific for a target antigen, for lack of written description. The reasoning applies to other receptor-based therapies.
- Ariad Pharmaceuticals, Inc. v. Eli Lilly & Co., 598 F.3d 1336 (Fed. Cir. 2010) (en banc), confirms that written description is a separate requirement from enablement.
- Amgen Inc. v. Sanofi, 598 U.S. 594 (2023), held that a patentee claiming a functionally defined genus must enable the full scope of the claim. Applied to TCR or receptor genera defined by binding function, this makes it hard to rely on a handful of examples.
- The enablement factors from In re Wands, 858 F.2d 731 (Fed. Cir. 1988), remain the working framework.
Practical implication: disclose a meaningful set of species, with sequences, plus structure-function data (for example, which CDRs, residues, or domains are responsible for activity). Include variants and fallback sequences at filing, because you cannot add them later.
Sequence-defined claims
Claims commonly use percent identity (“at least 90% identical to SEQ ID NO: 3”), conservative substitutions and consensus formulas. Each needs support in the description, so include enough species and data to justify the breadth claimed.
Eligibility under § 101
Engineered cells are generally eligible, following Diamond v. Chakrabarty, 447 U.S. 303 (1980), because they are human-made products with markedly different characteristics. Claims that capture naturally occurring cells or sequences face risk under Association for Molecular Pathology v. Myriad Genetics, Inc., 569 U.S. 576 (2013), which held natural DNA ineligible while allowing cDNA. In re Roslin Institute (Edinburgh), 750 F.3d 1333 (Fed. Cir. 2014), held cloned animals identical to their natural counterparts ineligible. For isolated TILs, Tregs, or MSCs, claims should highlight structural or functional differences from natural cells, such as engineered modifications, defined compositions, or altered characteristics.
Deposits
Where sequences cannot fully describe a biological material, such as a unique cell line, a hybridoma, or an iPSC line, a deposit may be required to satisfy enablement (37 C.F.R. §§ 1.801–1.809 and the Budapest Treaty). A deposit complements the sequence listing and does not replace it.
Part 8: Drafting Strategy
- Plan the sequence architecture early. List all sequences (receptors, linkers, tags, edits, vectors, primers) and assign identifiers before drafting, not after.
- Disclose fallback tiers. Include broad, intermediate and narrow sequence embodiments so claims can be narrowed during prosecution without new matter.
- Include variants generously. Variants, truncations and codon-optimized versions should be in the original filing if you may ever want to claim them.
- Tie sequences to data. Show that listed sequences work, for example through binding, cytotoxicity, persistence, or editing efficiency data.
- Use consistent nomenclature. State which CDR numbering scheme you use (IMGT, Kabat, or another) and use it consistently.
- Keep the listing synchronized with amendments. If you revise claims or add sequences during prosecution within the original disclosure, update all cross-references.
- Coordinate across the family. Maintain the same SEQ ID NO numbering across related filings where possible to reduce errors and simplify continuations.
- Treat the listing as a deliverable, not an afterthought. Assign a person to own the validation and cross-check step.
Part 9: International Perspective
European Patent Office. Under Rule 30 EPC, sequence listings must comply with ST.26 for applications filed on or after July 1, 2022. A listing filed after the filing date is not part of the description and the EPO may invite its submission with a late-furnishing fee. Substantively, Article 53(a) EPC and Rule 28(c) EPC bar inventions that use human embryos for industrial or commercial purposes, which matters for human embryonic stem cell claims (see Brüstle v. Greenpeace, CJEU C-34/10 and WARF, EPO Enlarged Board G 2/06). Methods of treatment are excluded under Article 53(c), so cell therapy claims are commonly drafted as product claims or in “for use” format.
China (CNIPA), Japan (JPO) and Korea (KIPO). Major offices moved to ST.26 alongside the PCT transition, but procedures for late filing, translation and conversion differ. Check local requirements before national-phase entry.
India. The Indian Patent Office has moved toward ST.26 for filings from the same transition date, but verify current practice and filing formalities. Substantive hurdles are significant for cell therapy: Section 3(i) excludes methods of medical treatment, Section 3(c) excludes discoveries of substances occurring in nature and Section 3(j) excludes plants and animals in whole or in part, other than microorganisms. Product claims to engineered cells with clear structural and sequence definition are more likely to pass than claims to natural cell populations or treatment methods. Section 3(d) may also be raised for new forms of known substances.
Filing strategy. Because a PCT application with an ST.26 listing serves as the foundation for national-phase entries, get the listing right once. Translation issues, especially for free-text qualifiers, should be planned for in advance.
Common Errors
| Error | Consequence | Fix |
| Omitting short linkers, tags and 2A peptides | Formal notice; sequences missing from disclosure | Scan the description and figures for every recited sequence above the threshold |
| Missing SEQ ID NOs in figures or claims | Notice to comply | Add identifiers everywhere the sequence appears |
| Citing accession numbers instead of sequences | Incorporation-by-reference and support problems | Disclose the sequence itself |
| Missing mandatory qualifiers in XML | Validation failure | Use the WIPO Sequence tool and fix all errors |
| Converting ST.25 to ST.26 without review | Discrepancies and possible new-matter issues | Compare converted output to the original and include the required statement |
| Listing the fusion but not its claimed domains | Weak claim support | List individual recited domains as separate sequences |
| Ignoring modified nucleotides in guides | Inaccurate or incomplete listing | Apply ST.26 rules for modified bases and annotate properly |
| Filing a listing late | No disclosure value in PCT; possible fees | File it complete with the application |
Common Misconceptions
- “Sequence listings only matter for antibodies and CARs.” Any recited sequence above the thresholds needs an identifier, including guide RNAs, primers and vector elements.
- “A gene name or accession number is enough for known sequences.” Not for material aspects of the invention. Database entries change and incorporation by reference is restricted.
- “I can add sequences later if I need them.” Not without adding new matter.
- “The listing is a formality that the vendor handles.” It is part of the disclosure and it needs attorney and scientist review.
- “Functional claim language will cover any receptor that works.” After Amgen and Juno, broad functional genus claims need structural support.
- “A deposit replaces the need for sequences.” A deposit complements the disclosure, but sequences remain necessary where they are recited.
Practical Checklist
Before drafting
- Inventory all sequence-bearing elements across the platform: receptors, editing reagents, vectors, assays.
- Identify which sequences may be claimed and which are only supporting.
- Decide on numbering scheme and nomenclature conventions.
During drafting
- Add SEQ ID NOs everywhere a qualifying sequence appears.
- Include tiered species and variants with supporting data.
- Avoid relying on accession numbers for essential sequences.
Before filing
- Generate the XML with the WIPO Sequence tool and validate it.
- Cross-check text, claims, figures and the listing.
- Confirm the incorporation-by-reference statement and file details.
- If converting from ST.25, review the output and include the required statements.
After filing
- Track notices and deadlines.
- Keep identifiers consistent across continuations and national-phase filings.
- Update the listing only within the bounds of the original disclosure.
Conclusion
As cell therapy moves beyond CAR-T, the sequence content of patent applications becomes broader and less obvious. TCR chains, guide RNAs, donor templates, vector elements and assay primers all count. The sequence listing is a substantive part of the disclosure, not a paperwork step and its completeness affects the breadth of claims you can pursue, how well they survive challenges and how smoothly your family moves through the PCT and national phases.
Companies that inventory sequences early, draft with tiered fallbacks, validate their ST.26 files and align disclosure with the case law on written description and enablement will be in the strongest position to secure durable protection across the growing range of cell therapy platforms.
Would you like me to put this article into a Word or Markdown file, or adapt it for a specific audience, such as a biotech IP newsletter or an Indian patent practice?