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

Bottom line: In selective autophagy, cargo receptors link a substrate
(glycogen, mitochondria, ER, bacteria, misfolded proteins) to ATG8-family
proteins on the forming autophagosome. GO has obsoleted the process
term GO:0061753 substrate localization to autophagosome because it
restates the receptor's molecular function (GO:0160247 autophagy cargo
adaptor activity), and it has no single replacement: each annotation should
move to the selective-autophagy process its paper supports. We pulled
every affected row from QuickGO (5 direct experimental annotations on
TOM1, Stbd1, IRGQ and SMURF1, 1 direct ISO row on Gabarapl1, 5 MF rows
that use the term in an extension, and 10 further ISS/ISO ortholog rows),
mapped each to a destination term, and found two errors in the upstream table.
The GO ontology ticket closed on 2026-08-10, QuickGO reports GO:0061753 as
obsolete as of 2026-10-04, and the annotation tracker is still open. Two
existing reviews here are now stale: human/RETREG2 and human/STBD1 each
ACCEPT an IEA row for GO:0061753 and list it in the strictly validated
core_functions, though both still pass against the repo's older ontology
cache.

The IRGQ case is the one open scientific question: no GO term covers
autophagy of misfolded MHC-I, so it is a candidate new-term request.

Overview

A GO obsoletion proposal retired 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 the pure queueing obsoletion trackers, two local reviews already carry
the obsolete term: genes/human/RETREG2 and genes/human/STBD1 each have a
GO:0061753 row and a core_functions use, where term ids are strictly
validated once the local ontology cache catches up.

Upstream tickets

Replacement-term status. GO did not mint a blanket replacement for this
obsoletion. GO:0061753 now resolves as obsolete in QuickGO, and the concrete
transfer destinations below already exist; IRGQ still exposes a narrower NTR
opportunity because no MHC-I-selective-autophagy term exists (verified
2026-08-08; GO:0061753 rechecked 2026-10-04):

GO id Label Aspect
GO:0061753 substrate localization to autophagosome (obsolete) 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) 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 — existing reviews go stale

genes/human/RETREG2/RETREG2-ai-review.yaml 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.

genes/human/STBD1/STBD1-ai-review.yaml has the same pattern for the
glycogen cargo receptor:

The biology is similarly clear: STBD1's receptor activity is already captured
as GO:0160247 autophagy cargo adaptor activity, and the selective-autophagy
process is GO:0061723 glycophagy. The obsolete generic BP is redundant.

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 upstream accessions has a review in this
repo (genes/human/TOM1, genes/human/IRGQ, genes/human/SMURF1,
genes/mouse/Stbd1, genes/mouse/Gabarapl1 all absent). Human STBD1 is
the reviewed orthologous glycophagy receptor and now a local stale row.

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.
  2. STBD1 (human, UniProt O95210) — genes/human/STBD1/. Already reviewed;
    the GO:0061753 IEA row and the core_functions entry should be reduced to
    the existing GO:0061723 glycophagy coverage.

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. Its
    human ortholog is now reviewed, and the mouse review still 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. Refresh RETREG2 and human STBD1 first. The obsoletion has landed
    upstream, and both reviews already carry redundant GO:0061753 rows in
    existing_annotations and core_functions. Drop the obsolete generic BP in
    favor of their existing reticulophagy and glycophagy terms.
  2. Then mouse Stbd1 as the clean direct transfer case, cross-checking
    against the existing human STBD1 and GAA reviews.
  3. 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.
  4. Then SMURF1 and TOM1; Gabarapl1 last.
  5. Report the two table corrections (missing direct Gabarapl1 row; its ISO/MGI
    provenance) back to go-annotation#6497.
  6. 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 upstream annotations are in play, but the
obsoletion has landed and two local human reviews now carry obsolete
GO:0061753 in the strictly validated core_functions slot. The IRGQ term gap
is also an actionable new-term contribution.

Status