Applying WIPO Standard ST.26 to modified, structured and multi-component RNA inventions

1. Introduction

The success of mRNA vaccines and the growth of RNA therapeutics (protein-replacement mRNA, self-amplifying RNA, circular RNA, mRNA-encoded antibodies, in vivo gene editing payloads and cancer neoantigen vaccines) have made RNA sequences central to patent filings. In these filings, the sequence listing is not an administrative afterthought. It is the formal, machine-readable part of the disclosure that defines each sequence you rely on in the specification and claims. Errors in it can trigger notices to comply, indefiniteness rejections, loss of searchability and, in the worst cases, argument over what was actually disclosed.

This article explains how the current sequence-listing standard applies to mRNA inventions, focusing on the issues that are particular to RNA: how to represent uracil, modified nucleotides, caps, poly(A) tails and untranslated regions, how to treat encoded proteins and codon-optimized variants and how to align the listing with the specification and the claims.

2. The Legal and Technical Framework

2.1 WIPO Standard ST.26

Since 1 July 2022, patent offices that follow WIPO standards, including the USPTO, EPO and the PCT receiving offices, require nucleotide and amino acid sequence listings to be filed in WIPO Standard ST.26, an XML-based format. It replaced the older text-based ST.25 for applications filed on or after that date. A listing in the old ST.25 format is not acceptable for new applications filed after the switch, even if a priority application used it, so converting priority-document listings is a real task for family members filed after July 2022.

2.2 U.S. rules

In the United States, the sequence-listing provisions are found in 37 C.F.R. ยงยง 1.831 through 1.835. In summary:

  • An application that discloses nucleotide sequences of ten or more specifically defined nucleotides, or amino acid sequences of four or more specifically defined amino acids, must include a sequence listing as a separate part of the disclosure, in compliance with ST.26.
  • The listing is filed as an XML file through Patent Center. It is not counted among the specification pages for page-count-based fees.
  • The specification must include an incorporation-by-reference paragraph identifying the file name, creation date and size in bytes of the XML file.
  • If a listing is submitted after the filing date, it must be accompanied by a statement that it includes no new matter.
  • Where a sequence is disclosed in the application in a way that falls under the rule (in the text, claims or drawings), the sequence must be identified by its sequence identifier (“SEQ ID NO:”) and should not be reproduced as a sequence string in the body of the specification, apart from the listing, except as the rule permits.

If the listing is missing or non-compliant, the Office issues a notice requiring compliance with a time limit and fees for extension of time may apply. These notices delay examination and for an application with time-critical claims, such as a vaccine candidate in an active outbreak or a Track One filing, delay has strategic cost.

2.3 Which sequences must be listed

The rule applies to sequences that are specifically defined at each position. A sequence of ten or more nucleotides that contains only “n” positions for modified bases is treated differently from one composed of defined residues. As a practical matter, you should list:

  • The full mRNA or RNA construct sequence and every distinct variant described or claimed.
  • Each 5′ UTR, 3′ UTR, signal peptide, coding region and poly(A) region that is disclosed as a distinct sequence of ten or more defined nucleotides.
  • The corresponding encoded amino acid sequences (of four or more residues), including antigens, linkers, epitope tags and engineered variants.
  • Primers, probes, guide RNAs, siRNAs, template DNA and plasmid inserts used in the invention, if they meet the length threshold.
  • Non-natural or modified sequences, subject to the special rules below.

Sequences that are shorter than the thresholds are not required to appear in the listing and may appear in the text, but be consistent. Many practitioners list borderline sequences anyway so that claims can refer to them by SEQ ID NO.

3. Representing RNA in ST.26

3.1 The molecule type and the uracil convention

This is the point where mRNA filers most often make mistakes. Under ST.26, nucleotide sequences are represented using the lowercase letters a, c, g and t, even for RNA. Uracil is represented as “t” and the molecule type is identified in the feature table by setting the /mol_type qualifier of the source feature to “mRNA,” “other RNA” or another appropriate value, as the standard allows. Do not enter “u” in the sequence. Do not enter uppercase letters. If you paste a sequence from a laboratory tool that outputs “AUGC,” it must be converted to lowercase and U must be changed to t.

The specification, on the other hand, can describe RNA in conventional terms (using U), but the specification should state that, in the sequence listing, uracil is represented as thymine, so as to make consistent the relationship between text and listing. A helpful sentence is: “In the sequence listing, RNA sequences are presented as DNA symbols with T representing U.” This avoids arguments that the listing sequence is DNA rather than RNA.

3.2 Why this matters for claims

A claim that recites “an mRNA comprising the sequence of SEQ ID NO: 1” is being read against a listing in which the sequence is expressed with t. In litigation or prosecution, any difference between the text and the listing can be exploited. Define the convention early and keep the claims neutral. Claims referencing “wherein the uracil of SEQ ID NO: 1 is N1-methylpseudouridine” are better than claims that try to embed the substitution in the sequence string.

3.3 Mandatory elements of the XML file

ElementWhat to supply
Document type and DTD versionThe root element ST26SequenceListing with the current DTD version.
Header informationSoftware name and version, production date, applicant file reference, application identification if known, earliest priority information, applicant name(s), invention title (one per language) and the total quantity of sequences.
SequenceData blocksOne block per sequence, with a unique sequenceIDNumber, an INSDSeq element including the length, molecule type (DNA, RNA or AA), division code, feature table and sequence residues.
Feature tableAt least a source feature with an /organism qualifier and an appropriate /mol_type qualifier. Add features for UTRs, coding regions and modified bases, as described below.
Language-dependent free textFree-text qualifiers (notes) must be in English and for languages in the application, an additional language version is also provided where required.

Do not write XML by hand. Use WIPO Sequence, the official authoring and validation tool, or another compliant tool and always run the validator before filing. Validation will catch many structural and syntactic errors, but it cannot tell whether your biology is right.

4. Modified Nucleotides, Caps and Tails

Therapeutic mRNA rarely consists only of natural A, C, G and U. A typical construct has N1-methylpseudouridine or pseudouridine in place of uridine, a 5′ cap, optimized UTRs, a coding region and a poly(A) tail and sometimes additional modifications such as 5-methylcytidine or specialized cap analogs. ST.26 offers specific mechanisms for these.

4.1 Modified bases

  • A modified nucleotide in the sequence is represented by the base letter that corresponds to its unmodified parent if the standard says so, or by “n” if it is not an acceptable representation.
  • Each modified residue should be described by a modified_base feature with the location and a /mod_base qualifier. The standard lists accepted abbreviations for common modified bases (the standard’s annex of modified nucleotides). If the modification is not on the list, use “OTHER” and include a /note qualifier that names the modification in words.
  • Where the whole molecule has uniform modification (for example every U is replaced with N1-methylpseudouridine), the ST.26 approach requires careful handling. Consult the current standard to determine whether each affected position needs an individual feature or whether a single, appropriately framed feature description is acceptable. Practice differs among offices, so confirm with counsel.
  • Treat “n” as an unspecified or unknown residue. A sequence in which most positions are “n” is not specifically defined and may fall below the threshold for listing.

The decision about how to represent a modification can have claim-scope consequences. If the listing represents N1-methylpseudouridine positions as plain “t,” a claim to “SEQ ID NO: 1” may be read to cover unmodified RNA. If the listing marks modifications, the claim may be read as narrower. Draft specification text that expressly explains the intended meaning and consider dependent claims that recite the modification separately.

4.2 Cap structures

The 5′ cap (for example Cap 0, Cap 1 or a co-transcriptional trinucleotide cap analog such as CleanCap-type structures) is a chemical structure that cannot be fully described in a sequence string. The sequence listing should represent the nucleotide sequence of the transcript. The cap itself should be described in the specification and, if it is claimed, in the claims, with chemical structure drawings or clear names. Where the standard permits, you may annotate the 5′ end with a feature such as a modified_base or misc_feature with an explanatory note, but do not assume the listing alone captures the cap. If the cap analog contributes nucleotides that count in the sequence, describe how they are numbered, to avoid positional confusion.

4.3 Poly(A) tails

A poly(A) tail is usually shown as a string of a’s. If the tail has ten or more specifically defined adenosines, it is a nucleotide sequence within the rule. Two practical issues arise:

  • Length variability: Actual tails are heterogeneous. If your specification describes a tail of “about 100” adenosines, the listing needs a defined representative sequence, such as a tail of 100 residues and the text should clarify that the listed length is exemplary.
  • Segmented tails: Some constructs use a tail interrupted by a linker or a non-adenosine segment. Represent it exactly and annotate it with a feature (for example misc_feature with a /note, or a more specific feature key where appropriate).

4.4 UTRs and coding regions

Annotate regulatory and coding regions in the feature table using standard keys such as 5’UTR, 3’UTR and CDS, with accurate location ranges. For a CDS feature, include the /translation qualifier where the standard requires it, so that the encoded protein is represented. Proteins encoded by the mRNA should be listed as separate amino acid sequences with their own SEQ ID NOs, if they have four or more specifically defined amino acids and the CDS feature should be consistent with them. Incorrect location numbers are an easy and common error, especially when an upstream sequence is lengthened or shortened late in drafting.

4.5 Other RNA formats

  • Self-amplifying RNA (saRNA/replicon): The nonstructural protein genes and the subgenomic promoter should be annotated and each encoded polyprotein or processed protein listed as appropriate. The sequences can be very long, so run the validator carefully for file size and format.
  • Circular RNA: ST.26 provides for indicating that a molecule is circular (the topology is flagged in the XML). Provide the linear sequence, indicate where the circle opens and describe the splicing elements.
  • trans-amplifying systems, tRNA, siRNA and guide RNA: Each distinct sequence of ten or more defined nucleotides needs its own entry. Chemical modifications such as 2′-O-methyl and phosphorothioate linkages are annotated with modified-base or misc_feature entries and described in text.
  • Template DNA and plasmids: If the application discloses the plasmid or DNA template, list the sequence as DNA, not RNA and note the relationship to the mRNA in the description.

5. Encoded Proteins, Codon Optimization and Variants

Vaccine claims often cover an antigen, a nucleic acid encoding it and compositions comprising that nucleic acid. The sequence listing should reflect each of those layers.

  • Antigen amino acid sequence: one SEQ ID NO for the wild-type antigen and one for each engineered variant (for example the stabilized prefusion spike with proline substitutions). Number mutations by reference to the SEQ ID NO that you list.
  • Wild-type coding sequence: if disclosed.
  • Codon-optimized coding sequences: each optimized version is its own sequence with its own identifier. A claim such as “a nucleic acid at least 90% identical to SEQ ID NO: 5 and encoding the polypeptide of SEQ ID NO: 2” requires that SEQ ID NO: 5 and SEQ ID NO: 2 both exist in the listing and be consistent.
  • Full-length mRNA: UTRs plus coding region plus tail, as the commercial sequence.
  • Signal peptides, linkers and tags: if they are four or more residues, include them.
  • Sequence variants and consensus sequences: if the specification discloses a consensus or a variable position, such as “X may be A or G,” use the standard’s ambiguity symbols or variant features rather than listing sets of sequences without explanation. Do not use “n” or “X” to hide uncertain residues if you intend to claim them.

Remember the relationship between claims and disclosure. After the U.S. Supreme Court’s 2023 decision in Amgen Inc. v. Sanofi, broad functional genus claims face substantial enablement scrutiny. In mRNA filings this means that listing and disclosing a wide range of working variants, not just a single lead sequence, can matter for supporting claims to sequence families defined by percent identity or function. The sequence listing should include the species you actually have and have tested, so the claims can be tied to concrete SEQ ID NOs.

6. Aligning the Listing With the Specification and Claims

6.1 Referring to sequences in the text

  • Every sequence that appears in the listing should be referred to by SEQ ID NO in the description and, where claimed, in the claims.
  • Where a sequence appears in a figure (for example an alignment), the figure legend should identify the sequences by SEQ ID NO.
  • If an alignment or table lists multiple sequences, each sequence of the required length must have its own listing entry, even if it is displayed in the figure as a fragment.
  • Avoid describing sequences in tables in the body of the specification when they should be in the listing. Repeating sequence strings in two places is an invitation to inconsistency.

6.2 Incorporation by reference

Include a short, accurate paragraph in the specification, such as: “The instant application contains a Sequence Listing which has been submitted in XML format and is hereby incorporated by reference in its entirety. Said XML file, created on [date], is named [filename] and is [size] bytes in size.” Update the file name, date and size every time you regenerate the XML, since the figures change with each edit.

6.3 Priority documents and later filings

  • If your priority application is a provisional with a legacy ST.25 listing or sequences embedded in the text, create a compliant ST.26 listing for the non-provisional and make sure no new matter is introduced. If the content of the listing changes in substance, you may lose priority for the changed content.
  • For PCT applications, the ST.26 listing is filed as part of the international application. Do not rely on the listing from the priority filing to satisfy a requirement in the international phase.
  • For national-phase entries, check each office for specific rules about whether the listing must be refiled, translated or reformatted.

7. Common Errors in mRNA Listings

ErrorConsequenceHow to avoid it
Using “u” or uppercase letters in the sequenceValidation failure or an incorrect, DNA-style representation.Convert to lowercase and replace U with t before import. Use the WIPO tool for validation.
Mislabeling mol_type (DNA instead of RNA, or the reverse)Ambiguity over whether the claimed molecule is DNA or RNA.Set the molecule type deliberately for each entry; plasmids are DNA, transcripts are RNA.
Failing to list encoded proteinsClaims to the protein have no listing support; missing SEQ ID NOs in claims.List each protein of four or more residues and link it to the CDS.
Wrong feature locations after sequence editsAnnotations point to wrong region; examiner questions disclosure.Re-verify the feature table after any change and compare to a map of the construct.
Misrepresenting modified basesClaim scope unclear or listing rejected.Use modified_base features and /mod_base or /note. Explain conventions in the specification.
Inconsistent sequence between text, claims and listingRejections for indefiniteness or lack of written description.Treat the listing as the single source of truth and generate text excerpts from it, not the reverse.
Sequence number gaps, duplicate SEQ ID NOs or incorrect total quantityValidation errors.Regenerate numbering after deletions; confirm sequenceTotalQuantity.
Incorrect incorporation-by-reference dataNotice to comply.Update file name, date and size after each export.
Listing unspecified sequences with long “n” runsSequence is not a “specifically defined” sequence; the listing may be unnecessary or misleading.Decide whether to omit or describe the sequence in words.
Leaving out variants disclosed in tables or figuresNotice to comply or lost basis for claims.Search the specification for every sequence of the threshold length.

8. A Practical Workflow

  1. Collect the sequences. Get the electronic source files from the lab, not retyped copies. Confirm which sequences are RNA, DNA and protein.
  2. Define conventions. Decide how to represent modifications, caps and tails. Draft the description paragraph that explains these conventions.
  3. Build the listing. Use WIPO Sequence. Enter the sequence, set the molecule type, add features (source, 5’UTR, CDS, 3’UTR, modified_base, misc_feature) and add the encoded proteins as separate entries.
  4. Validate. Run the validator, resolve all errors and read every warning. Treat warnings as a to-do list.
  5. Cross-check biology. Have a scientist verify that the translation of each CDS matches the listed protein, that mutations are at the stated positions and that UTR and CDS boundaries match the plasmid map.
  6. Cross-check the text. Search the specification and claims for every SEQ ID NO and verify the description. Confirm all sequences of required length are in the listing.
  7. Update the incorporation-by-reference paragraph with the final file details.
  8. File and preserve. Upload the XML file as the sequence listing document type in Patent Center. Keep a copy of the final XML and validation report for your records.
  9. Re-run the process for every amendment that changes any sequence or the order of SEQ ID NOs.

9. Strategic Considerations for Vaccine and Therapeutic Applicants

  • Provisionals: Even where a formal listing is not required in a provisional, include the actual sequences in the provisional text or an accompanying listing. A later claim to a sequence not disclosed in the provisional cannot claim its priority.
  • Fast filings in outbreaks: In a rapidly evolving situation, do not let sequence-listing formalities delay a filing. File first with sequences in the application, then correct the formal listing, but be mindful that submissions after filing must include the no-new-matter statement and cannot add content.
  • Claim structure: Layer claims across the antigen, the nucleic acid (with percent identity ranges), the full construct (including modifications), the lipid nanoparticle formulation and methods of use or manufacture. Each layer depends on correct sequence references.
  • Freedom to operate: The listing and its XML form are searchable in public sequence databases and in commercial patent-sequence search tools. Accurate, well-annotated listings make your applications discoverable and equally make competitors’ applications easier for you to analyze when doing clearance work.
  • Trade secrets and sequence disclosure: Patent filing requires public disclosure of the sequence (subject to the usual publication timeline). Consider which optimized sequences to disclose and which to retain as know-how, noting the enablement and best-mode implications.
  • Consistency across a family: Maintain a master sequence table that maps each SEQ ID NO across family members, so that continuations, divisionals and foreign counterparts remain aligned.

10. Conclusion

For mRNA vaccines and therapeutics, the sequence listing is a legal document that carries real weight. Under ST.26, RNA is entered with lowercase DNA symbols, uracil as t and the molecule type specified separately; modifications, caps and tails require careful annotation and explanatory text; and every encoded protein and every variant you intend to claim must appear with an identifier. Build the listing with the official tool, validate it, verify the biology independently, align the text and claims to it and keep the incorporation-by-reference details current. A disciplined approach protects your priority date, avoids avoidable office actions and supports the claim scope that your science deserves.

Leave a Reply

Your email address will not be published. Required fields are marked *