GPD (Pichia kudriavzevii / Issatchenkia orientalis, UniProtKB A8W1U0) notes

No deep-research provider was available in this environment. This file is a manual
literature and record review.

Identity

Relationship to the reference proteome

A8W1U0 is not in a reference proteome. A UniProtKB REST search for every EC 1.1.1.8
entry of NCBITaxon:4909 returns exactly two records (2026-10-28): A8W1U0 (388 aa, no
proteome) and A0A2U9RB90 (415 aa, UP000249293: Chromosome 5, ORF C5L36_0E02100).
A8W1U0's own record cross-references FungiDB:C5L36_0E02100, i.e. the same locus.
Aligned, A8W1U0 corresponds to residues 28-415 of A0A2U9RB90 with 385/388 identities
(99.2%; R89K, F165S, R228H in A8W1U0 numbering); A0A2U9RB90 carries a 27-residue
N-terminal extension (MFLISKSLHFTRSFSNTIKTLNRSNIK) that the A8W1U0 gene model
starts after. These are different strains (an industrial glycerol producer versus the
sequenced reference isolate) and the shorter model is the one with functional data
behind it, since the 388-residue product complements S. cerevisiae gpd1Δ. The
relationship is recorded under alternative_records as
REFERENCE_PROTEOME_EQUIVALENT, with the gene-model difference stated. Which start
codon is used in vivo is unresolved and is raised in suggested_questions; the review
does not assert that either model is defective.

Function

Review decisions (10 IEA rows, no PENDING left)

Term Action
GO:0141152 glycerol-3-phosphate dehydrogenase (NAD+) activity ACCEPT (core)
GO:0047952 G3P dehydrogenase [NAD(P)+] activity MODIFY → GO:0141152
GO:0016616 oxidoreductase activity, CH-OH donors, NAD(P) acceptor MODIFY → GO:0141152
GO:0051287 NAD binding KEEP_AS_NON_CORE
GO:0042803 protein homodimerization activity KEEP_AS_NON_CORE
GO:0006072 glycerol-3-phosphate metabolic process MODIFY → GO:0046167
GO:0046168 glycerol-3-phosphate catabolic process MODIFY → GO:0046167 (+ source defect)
GO:0005975 carbohydrate metabolic process MODIFY → GO:0006114
GO:0005829 cytosol ACCEPT
GO:0005634 nucleus KEEP_AS_NON_CORE

No NEW rows were added. Glycerol biosynthesis and glycerol-3-phosphate biosynthesis
both enter the review as MODIFY replacements of terms the gene already carries, which
is the refinement path; neither needed the NEW bar. No process term was proposed for
oxalic acid production: GPD is deleted in those strains, and a gene whose removal
improves a pathway is not a participant in it.

Why GO:0006973 intracellular accumulation of glycerol was not added

GO:0006973 intracellular accumulation of glycerol was considered as a NEW row and
declined. The participation test is not the obstacle: an enzyme that makes the glycerol
that accumulates does part of the work of accumulating it, and genes/yeast/GPD1
carries the term as a core process. The obstacle is whose phenotype was measured. The
only accumulation data for this protein come from expressing it in a heterologous
S. cerevisiae host, where the osmotic response machinery, the glycerol channel Fps1
and the transcriptional regulation are all S. cerevisiae's, not
P. kudriavzevii's PMID:18452539. No osmotic phenotype, and no intracellular glycerol measurement, has been
reported for P. kudriavzevii itself: the only native-host data are the fermentation
knockout, which measures glycerol in the broth, not retained inside the cell
PMID:41617712. Accumulation is a
process of the cell that retains the solute, so transferring it on the strength of a
heterologous host's regulation would be asserting more than was measured. If an osmotic
or intracellular-glycerol phenotype is reported for P. kudriavzevii, the term should
be added; suggested_experiments item 2, which now scores localisation under
hyperosmotic conditions, is the natural place for that measurement.

Cross-review follow-up (not in this PR)

GO:0006072 glycerol-3-phosphate metabolic process is MODIFY here but ACCEPT in
the merged genes/yeast/GPD1 review, from the same PAINT node PTN000198006. Per
CLAUDE.md's "correct but general terms are MODIFY" rule the call in this review is the
one to keep, since glycerol 3-phosphate is this enzyme's product and the direction-
specific descendant GO:0046167 is supported. Aligning genes/yeast/GPD1 is a
follow-up on that review, outside this change.

CMMMech cross-check (data/records/allanite-leachate.yaml)

The grounding of the gpd_glycerol3P_dehydrogenase node to UniProtKB:A8W1U0 is
correct in species, gene and enzyme, and the InterPro/Pfam/EC fields match the UniProt
record exactly. Three points for the CMMMech curator (not edited here; CMMMech is
read-only for this work):

  1. functional_role is self-contradictory on JL04. It says GPD is "deleted in
    Issatchenkia orientalis (Pichia kudriavzevii) strain JL02 (and retained in JL04)",
    while the accompanying evidence item in the same block says Supplementary Table 1
    documents "the Δgpd knockout genotype in engineered strains JL02, JL03, and JL04".
    The latter is right: JL04 descends from JL02, so GPD is deleted, not retained, in
    JL04. The "(and retained in JL04)" phrasing appears to have been copied from the
    neighbouring PYK1 and PYC entries, where retention in JL04 is the correct claim.
  2. No locus or reference-proteome field is given for this protein example, unlike
    the PYC entry. A8W1U0 is not in any reference proteome; the reference-proteome
    representation of the same locus is A0A2U9RB90 (UP000249293, ORF
    C5L36_0E02100). Recording that would make the grounding checkable. (The PYC
    entry's reference_proteome_id: UP000249646 is wrong and was reported from the
    PYC review; the GPD entry has no such field, so the error does not recur here.)
  3. The 14.72 g/L titer is quoted correctly but attached to the wrong one of two
    fermentations.     The paper reports two JL02 titers in the same paragraph, both in the
    main text: 14.7 g/L for the bioreactor comparison shown in Fig. 2c
    PMID:41617712, and 14.72 g/L for a separate
    fed-batch run PMID:41617712. CMMMech's evidence explanation pairs
    14.72 with "Fig. 2c", where the figure's number is 14.7. The titer itself is accurately
    quoted from the main text; only the fermentation it is attributed to would need adjusting,
    either by citing the SC-URA glucose-feeding run or by using 14.7 for Fig. 2c. Minor.

Nothing in CMMMech overstates this protein's function: calling it an "endogenous
NAD-dependent glycerol-3-phosphate dehydrogenase" whose deletion removes glycerol
byproduct is exactly what the paper shows, and the record does not claim any role for
GPD in oxalate formation or REE recovery.