Substrate Localization to Autophagosome (GO:0061753) — Obsoletion & Transfer

SCOPING OBSOLETION

Species: human, mouse

Genes: TOM1 IRGQ SMURF1 STBD1 GABARAPL1 RETREG2

Substrate Localization to Autophagosome (GO:0061753) — Obsoletion & Transfer

Overview

A GO obsoletion proposal will retire the biological-process term
GO:0061753 substrate localization to autophagosome — "The localization
process by which an autophagic substrate is delivered to a forming
autophagosome."

The ontology ticket is labelled MF_in_BP: the term describes what an
autophagy cargo receptor/adaptor does (a binding activity, now covered by
GO:0160247 autophagy cargo adaptor activity), not a distinct biological
process. There is no single replacement BP. Instead each annotation must be
transferred to the specific selective-autophagy process that the cited
paper actually supports (glycophagy, mitophagy, aggrephagy, xenophagy,
autophagosome-lysosome fusion, …), which makes this a per-annotation curation
job rather than a mechanical relabel.

Unlike most obsoletion trackers in this repo, one affected review already
exists here — genes/human/RETREG2 carries a GO:0061753 row and uses the
term in core_functions, where term ids are strictly validated. That review
breaks the moment the obsoletion lands.

Upstream tickets

Replacement-term status. No new term needs minting. Every proposed
destination already exists and resolves in OLS (verified 2026-08-08):

GO id Label Aspect
GO:0061753 substrate localization to autophagosome (to be obsoleted; still active) BP
GO:0061723 glycophagy BP
GO:0061909 autophagosome-lysosome fusion BP
GO:0061734 type 2 mitophagy BP
GO:0098792 xenophagy BP
GO:0035973 aggrephagy BP
GO:0016236 macroautophagy BP
GO:0006515 protein quality control for misfolded or incompletely synthesized proteins BP
GO:0160247 autophagy cargo adaptor activity MF
GO:0140580 mitochondrion autophagosome adaptor activity MF
GO:0140506 endoplasmic reticulum-autophagosome adaptor activity MF
GO:0038024 cargo receptor activity MF
GO:0043495 protein-membrane adaptor activity MF
GO:0030674 protein-macromolecule adaptor activity MF

Affected annotations (independently verified via QuickGO, 2026-08-08)

All UniProt accessions below were confirmed against the UniProt REST API; the
annotation rows were pulled from QuickGO directly rather than taken from the
upstream issue text.

(A) Direct experimental annotations to GO:0061753

Gene product Species UniProt Ev Reference Extension Upstream recommendation
TOM1 human O60784 IMP PMID:23023224 — → GO:0061909 autophagosome-lysosome fusion (± GO:0035973 aggrephagy)
Stbd1 mouse Q8C7E7 IMP PMID:20810658 part_of(GO:0061723) → GO:0061723 glycophagy; drop now-redundant extension
Stbd1 mouse Q8C7E7 IMP PMID:27358407 — → GO:0061723 glycophagy
IRGQ human Q8WZA9 IDA PMID:39481378 — → GO:0006515 + GO:0016236 (no MHC-I-selective-autophagy term exists)
SMURF1 human Q9HCE7 IMP PMID:22020285 has_target_start_location(GO:0005739), part_of(GO:0061734) → GO:0061734 type 2 mitophagy, plus GO:0098792 xenophagy from the same paper

(B) Experimental MF annotations carrying GO:0061753 in an extension

Gene product Species UniProt MF term Ev Reference Recommendation
Stbd1 mouse Q8C7E7 GO:0038024 IDA PMID:20810658 drop part_of(GO:0061753); keep part_of(GO:0061723)
Stbd1 mouse Q8C7E7 GO:0038024 IMP PMID:27358407 swap part_of(GO:0061753) → part_of(GO:0061723)
Gabarapl1 mouse Q8R3R8 GO:0043495 ISO PMID:21893048 swap part_of(GO:0061753) → part_of(GO:0061723)
IRGQ human Q8WZA9 GO:0030674 IDA PMID:39481378 drop part_of(GO:0061753); part_of(GO:0006515) already present
IRGQ human Q8WZA9 GO:0030674 IDA PMID:39481378 as above (differs only in has_input)

Two corrections to the upstream issue table

Worth flagging back to go-annotation#6497 before curators act on it:

  1. Gabarapl1 Q8R3R8 also has a direct GO:0061753 annotation (involved_in,
    ISO, PMID:21893048, assigned by MGI) that the issue's section (A) does not
    list — it only appears there as an MF-extension row. It needs a transfer
    decision too (presumably GO:0061723 glycophagy, matching the same paper).
  2. The Gabarapl1 MF row is ISO assigned by MGI, not an experimental
    GO Central annotation; the issue's "Impacted groups" table attributes it to
    "GO Central / mouse".

(C) Inferred rows that follow automatically

Ten further manual-but-non-experimental rows (ISS GO_REF:0000024, ISO
GO_REF:0000119 / GO_REF:0000121) exist on orthologs and will follow the
experimental transfers: rat Smurf1 (A0A0G2K612), rat Irgq (A0A8I5ZUT2),
rat Tom1 (A0ACM8Q5W0), rat Stbd1 (Q5FVN1), chicken TOM1 (O12940), mouse
Tom1 (O88746, ISS + ISO), mouse Irgq (Q8VIM9, ISS + ISO), mouse Smurf1
(Q9CUN6, ISO). Thousands of IEA rows (GO_REF:0000107 Ensembl,
GO_REF:0000108 GOC inter-ontology) propagate automatically and are out of
scope, per the upstream issue.

Impact on this repo

Tier 1 — an existing review goes stale

genes/human/RETREG2/RETREG2-ai-review.yaml is the only review in the repo
touching GO:0061753, and it is affected twice:

The core_functions occurrence is the hard problem: per CLAUDE.md, GOA-sourced
existing_annotations[].term.id values are not hard-validated, but
core_functions term ids are. Once GO:0061753 is obsoleted,
just validate human RETREG2 should be expected to flag the core_functions
entry. The fix is straightforward on the biology — RETREG2/FAM134A is an ER-phagy
receptor, so the reticulophagy BP (GO:0061709, already present) plus the MF
GO:0140506 endoplasmic reticulum-autophagosome adaptor activity (also already
on that review) carry the content; the GO:0061753 entry is redundant rather than
wrong and should simply be dropped.

The upstream issue notes that MF adaptor classes under GO:0160247 need
re-parenting; that work is go-ontology#31866, which is already closed. Repo
reviews carrying those MF terms should be spot-checked once the obsoletion lands
but are not expected to change:

Tier 3 — destination terms already used here

Reviews already annotating the transfer destinations give useful precedent for
how these BPs are used in this repo: GO:0061723 glycophagy —
human/GAA (two rows, both KEEP_AS_NON_CORE), human/ATG2A, human/ATG2B,
SCHPO/atg2, worm/atg-18. human/GAA is directly relevant: PMID:27358407
(the Stbd1/GAA double-knockout paper behind transfer A3) is the same
experimental system.

None of the five directly affected gene products has a review in this repo
(genes/human/TOM1, genes/human/IRGQ, genes/human/SMURF1,
genes/mouse/Stbd1, genes/mouse/Gabarapl1 all absent).

Scope

Candidate genes for initial review

Confirm accessions with just fetch-gene <organism> <gene> before starting.

Tier 1 — refresh required

  1. RETREG2 (human, UniProt Q8NC44) — genes/human/RETREG2/. Already
    reviewed; both the GO:0061753 existing_annotations row and the
    core_functions entry need revisiting. Highest priority: an existing review
    goes stale, and core_functions validation is strict.

Tier 2 — directly affected, not yet in repo

  1. STBD1 (mouse Q8C7E7; human ortholog O95210) — the cleanest case: two
    independent IMP papers, an unambiguous destination (GO:0061723 glycophagy),
    and an MF cargo-receptor annotation whose extension needs the same swap.
    Pairs naturally with the existing human/GAA review.
  2. IRGQ (human Q8WZA9) — the hard case. PMID:39481378 shows IRGQ routing
    misfolded MHC-I to lysosomal degradation via GABARAPL2/LC3B. No suitable
    selective-autophagy child exists, so this is a genuine
    proposed_new_terms opportunity rather than a transfer.
  3. SMURF1 (human Q9HCE7) — PMID:22020285 supports two destinations from
    one paper (GO:0061734 type 2 mitophagy and GO:0098792 xenophagy); a good test
    of one-annotation-splits-into-two.
  4. TOM1 (human O60784) — PMID:23023224; the recommendation moves the
    annotation from a cargo-sequestration BP to a fusion BP (GO:0061909), a
    different step of the pathway. Worth checking whether the myosin VI/TOM1
    evidence really supports fusion rather than delivery.
  5. GABARAPL1 (mouse Q8R3R8; human ortholog Q9H0R8) — an ATG8-family
    protein annotated as a GO:0043495 adaptor. Both its direct ISO row and its
    MF extension point at GO:0061753. Lower priority (ISO, not experimental) but
    it is the only ATG8-family member in the affected set.

Proposed approach

  1. Wait for the obsoletion to land. As of 2026-08-08 GO:0061753 is still
    active in OLS (is_obsolete: false), though go-ontology#32304 is labelled
    ready, so this could move soon.
  2. Refresh RETREG2 first, ideally before the obsoletion, since the fix
    (dropping a redundant core_functions entry) does not depend on the final
    destination chosen upstream. Regenerate with
    just fetch-gene human RETREG2, re-review the GO:0061753 row, then
    just validate human RETREG2.
  3. Then STBD1 as the clean transfer case, cross-checking against the
    existing human/GAA review.
  4. Then IRGQ, and use it to draft a proposed_new_terms entry for MHC-I
    quality-control autophagy — the concrete deliverable this project can send
    back upstream.
  5. Then SMURF1 and TOM1; GABARAPL1 last.
  6. Report the two table corrections (missing direct Gabarapl1 row; its ISO/MGI
    provenance) back to go-annotation#6497.
  7. Cross-reference the sibling obsoletion trackers in this repo — the
    ER exit site,
    vesicle targeting and
    synaptic vesicle docking pages —
    which share the same MF_in_BP "this BP is really a molecular function"
    rationale.

Priority

Medium. Only five experimental annotations are in play and none of those
gene products is reviewed here yet, so the immediate blast radius is small. It
ranks above the pure queueing trackers because (a) RETREG2 is an existing
review that breaks on obsoletion, in the strictly-validated core_functions
slot, and (b) the IRGQ term gap is an actionable new-term contribution.
Nothing is broken until the obsoletion is applied.

Status