Geranylgeranyl Reductase Activity — Obsoletion & Replacement

SCOPING OBSOLETION

Species: human, ARATH, TOBAC

Geranylgeranyl Reductase Activity — Obsoletion & Replacement

Bottom line: GO has obsoleted GO:0045550 geranylgeranyl reductase activity,
which had no enzyme cross-reference, and replaced it with GO:0102067
geranylgeranyl diphosphate reductase activity (EC 1.3.1.83; RHEA:26229). The
new term captures the plant/cyanobacterial CHLP reaction that reduces
geranylgeranyl diphosphate to phytyl diphosphate and can also reduce
geranylgeranyl-chlorophyll a. The open work is now the annotation and mapping
cleanup in #363:
Arabidopsis CHLP is the clean experimental replacement case, tobacco CHLP is in
the same family but has only NAS evidence, and human AKR1C3/AKR1B10 both come
from one aldo-keto reductase substrate-screening paper that needs review before
any move to GO:0102067. The archaeal IPR023590 mapping remains the suspicious
mapping because that family reduces lipid-linked geranylgeranyl chains rather
than free GGDP. As of 2026-10-04, none of the four affected gene products has a
review in this repo.

Overview

The GO obsoletion has replaced GO:0045550 geranylgeranyl reductase activity
with GO:0102067 geranylgeranyl diphosphate reductase activity. The replacement
narrows the substrate from generic "geranylgeranyl-X" to the diphosphate (GGDP),
aligning the term with the chlorophyll/tocopherol biosynthesis enzyme
(plant/cyanobacterial CHLP / ChlP / GGR) that catalyses the reduction of
geranylgeranyl-(diphosphate or chlorophyll) to phytyl form.

This project tracks the four experimental annotations on the upstream list and
queues per-gene reviews where useful — particularly the two human aldo-keto
reductase entries (AKR1C3, AKR1B10), which are likely a misuse of the
geranylgeranyl reductase term derived from in vitro substrate-screening data.

Upstream tickets

Obsoletion outcome

Obsoleted term ID Replacement
geranylgeranyl reductase activity GO:0045550 GO:0102067 geranylgeranyl diphosphate reductase activity

The current GO ontology cache on 2026-10-04 lists GO:0045550 as obsolete,
with replaced_by: GO:0102067; the live replacement carries EC 1.3.1.83 and
RHEA:26229 cross-references.

The replacement term is substrate-specific (geranylgeranyl-diphosphate),
so any annotation that was made because the enzyme reduces a different
geranylgeranyl-conjugated substrate (e.g. free geranylgeraniol, geranylgeranyl
attached to a chlorophyll/tocopherol intermediate, or geranylgeranyl-CoA) is
not automatically a fit for the replacement and should be considered for
removal or remapping rather than a mechanical move.

Affected experimental annotations (upstream list)

# Source Accession Symbol Taxon PMID Evidence Notes
1 TIGR AGI_LocusCode:AT1G74470 CHLP NCBITaxon:3702 (Arabidopsis thaliana) PMID:9492312 IDA Plant chlorophyll synthesis GGR — strong candidate to move cleanly to GO:0102067
2 UniProt UniProtKB:Q9ZS34 CHLP NCBITaxon:4097 (Nicotiana tabacum) PMID:10398704 NAS Tobacco chloroplastic GGR in the plant CHLP/GGR family; NAS evidence is weaker and may need removal rather than remapping if no traceable primary support is found
3 UniProt UniProtKB:P42330 AKR1C3 NCBITaxon:9606 (Homo sapiens) PMID:21187079 IDA Aldo-keto reductase; geranylgeranyl reductase claim likely in-vitro side activity — flag for review
4 UniProt UniProtKB:O60218 AKR1B10 NCBITaxon:9606 (Homo sapiens) PMID:21187079 IDA Aldo-keto reductase; same paper as AKR1C3 — flag for review

Group impact tally (from upstream): TIGR 1, UniProt 3.

Mappings flagged by upstream

The plant/cyanobacterial CHLP/GGR family in IPR010253/IPR011774 is
substrate-correct for the replacement; follow-up should verify that those
signatures and UniRule UR001995838 have landed on GO:0102067. IPR023590
(Digeranylgeranylglycerophospholipid reductase, archaeal lipid biosynthesis)
acts on a digeranylgeranylglycerophospholipid substrate that is not free GGDP,
so the IPR023590 mapping should be reviewed for whether GO:0102067 is the right
replacement.

Impact on this repo

None of the four affected genes currently have an *-ai-review.yaml in this
repo (verified via find genes -iname '*AKR1C3*' -o -iname '*AKR1B10*' -o -iname '*CHLP*' on 2026-10-04). This project is therefore primarily a
queueing exercise rather than a re-review of existing files.

Scope

Candidate genes for initial review

Listed in priority order. Each should be set up with
just fetch-gene <organism> <gene> before review begins.

  1. AKR1C3 (human, P42330) — Aldo-keto reductase 1C3. The PMID:21187079
    IDA annotation to GO:0045550 should be evaluated against the actual
    substrate(s) tested. AKR1C3's well-characterised functions are
    prostaglandin / steroid 17-keto / 3α-hydroxysteroid reductase activities;
    any geranylgeranyl-reductase activity is likely in-vitro screening, not
    the in-vivo function. Likely outcome: REMOVE or MARK_AS_OVER_ANNOTATED on
    the GO:0045550 entry.
  2. AKR1B10 (human, O60218) — Aldo-keto reductase 1B10. Same paper
    (PMID:21187079) and same critique as AKR1C3. Characterised activities
    include retinal reductase (all-trans-retinal to retinol) and farnesal
    reductase; the primary in-vivo function remains debated. Likely outcome
    similar to AKR1C3.
  3. CHLP (Arabidopsis, AT1G74470) — Geranylgeranyl reductase in
    chlorophyll/tocopherol biosynthesis. PMID:9492312 IDA annotation. This is
    the canonical use of the term and the cleanest candidate to move to
    GO:0102067; useful as a positive control for the replacement.
  4. CHLP (Nicotiana tabacum, taxon 4097, Q9ZS34) — Tobacco chloroplastic
    GGR in the plant CHLP/GGR family. UniProt REST confirms the taxon and
    protein name. PMID:10398704 NAS evidence is lower-strength; absent
    traceable primary support, this annotation may need removal rather than
    automatic remapping.

Implementation status

Open follow-up is tracked in
#363.

  1. Prioritise the human AKR pair. AKR1C3 and AKR1B10 are the most
    scientifically interesting cases here — both are well-studied human
    enzymes whose former IDA annotation to GO:0045550 deserves a critical
    evidence review rather than a mechanical remap. Start with AKR1C3.
  2. Use the plant CHLP as a positive control. Reviewing the Arabidopsis
    CHLP entry is straightforward and helps confirm that GO:0102067 is the
    right replacement for genuine in-vivo GGR enzymes.
    Treat the tobacco Q9ZS34 NAS annotation separately because NAS evidence
    may not provide enough traceable support for either ACCEPT or MODIFY.
  3. Verify the family mappings after ontology refresh. UR001995838,
    IPR010253, and IPR011774 should point to GO:0102067; IPR023590 should not
    be moved mechanically if its archaeal lipid substrate falls outside the
    term's substrate scope.
  4. Flag IPR023590 mapping question to upstream. The
    digeranylgeranylglycerophospholipid reductase family does not act on free
    GGDP, so noting this on the upstream issue (or via a comment on
    #6394) is a useful contribution even before any review work.

Priority

Medium — the human AKR1C3 / AKR1B10 entries are interesting reviews on their
own merits, independent of the obsoletion timing. The Arabidopsis CHLP entry is
a clean positive control for the replacement term; the tobacco CHLP entry is a
lower-priority NAS row that still needs traceable support.

Status