NAD(+)-dependent glycerol-3-phosphate dehydrogenase (EC 1.1.1.8) of the acid- and osmotolerant yeast Pichia kudriavzevii (Issatchenkia orientalis), a 388-residue cytosolic enzyme of the NAD-dependent glycerol-3-phosphate dehydrogenase family (PANTHER PTHR11728, NCBIfam TIGR03376, PIRSF000114) with an N-terminal Rossmann-fold NAD(+)-binding domain and a C-terminal substrate-binding domain carrying the conserved proton-acceptor lysine. Although the EC reaction is written as the oxidation of sn-glycerol 3-phosphate, the physiological direction in yeast is reductive: the enzyme reduces dihydroxyacetone phosphate from glycolysis to glycerol 3-phosphate using NADH, which glycerol-3-phosphate phosphatase then dephosphorylates to glycerol. It is therefore the committed and rate-controlling step of glycerol synthesis, which serves two purposes in yeast: accumulating glycerol as a compatible solute during osmotic stress, and reoxidising surplus cytosolic NADH. The protein complements Saccharomyces cerevisiae gpd1 and gpd1 gpd2 mutants, restoring osmotic tolerance and intracellular glycerol accumulation, and deleting the gene in I. orientalis abolishes glycerol as a fermentation byproduct.
The preferred accession for this review is A8W1U0. These relationships are manually assessed.
REFERENCE PROTEOME EQUIVALENT · Reviewed 2026-10-28
Reference proteome: UP000249293
A8W1U0 derives from a single genomic clone (EMBL EU186536) of an industrial glycerol-producing P. kudriavzevii isolate and belongs to no reference proteome. A0A2U9RB90 is the same gene in the sequenced reference isolate: both records cross-reference the same FungiDB locus C5L36_0E02100, and a UniProtKB search for EC 1.1.1.8 in NCBITaxon:4909 returns these two records and no others, so there is no competing paralog to confuse them. The sequences are not identical, and the differences are stated rather than glossed: A8W1U0 aligns to residues 28-415 of A0A2U9RB90 with 385/388 identities (99.2%; three substitutions, R89K, F165S and R228H in A8W1U0 numbering, consistent with two strains), and A0A2U9RB90's gene model adds 27 N-terminal residues that the A8W1U0 model starts after. Which initiator is used in vivo is unresolved, and no defect is asserted against either record: the shorter form is the one with functional data behind it, since the 388-residue product complements S. cerevisiae gpd1. This relationship transfers no GO annotation, and the review keeps A8W1U0 as its accession.
A0A2U9RB90 is a current member of the P. kudriavzevii reference proteome UP000249293, on chromosome 5 (UniProtKB REST record, retrieved 2026-10-28).
file:PICKU/GPD/GPD-record-evidence.yaml
DR Proteomes; UP000249293; Chromosome 5.
Both records carry the same VEuPathDB/FungiDB locus cross-reference, C5L36_0E02100, which is also A0A2U9RB90's ORF name. This is the evidence that the two records represent the same gene rather than two similar genes.
file:PICKU/GPD/GPD-uniprot.txt
DR VEuPathDB; FungiDB:C5L36_0E02100; -.
file:PICKU/GPD/GPD-record-evidence.yaml
DR VEuPathDB; FungiDB:C5L36_0E02100; -.
A0A2U9RB90 is 415 residues to A8W1U0's 388. Aligning A8W1U0 to residues 28-415 of A0A2U9RB90 gives 385/388 identities (99.2%) with three substitutions; the 27-residue N-terminal extension MFLISKSLHFTRSFSNTIKTLNRSNIK has no counterpart in A8W1U0. Identity was computed over the ungapped alignment of the two record sequences; the length difference is a gene-model difference at the 5' end, not a truncation that either record asserts.
file:PICKU/GPD/GPD-record-evidence.yaml
SQ SEQUENCE 415 AA; 45822 MW;
file:PICKU/GPD/GPD-uniprot.txt
SQ SEQUENCE 388 AA; 42762 MW;
A UniProtKB REST search for every EC 1.1.1.8 entry of NCBITaxon:4909 (retrieved 2026-10-28) returned exactly two records, A0A2U9RB90 and A8W1U0. The search is bounded by UniProt's EC annotation, so a family member lacking an EC number would not appear; no such entry was found by family search either.
file:PICKU/GPD/GPD-record-evidence.yaml
A0A2U9RB90 A0A2U9RB90_PICKU Glycerol-3-phosphate dehydrogenase [NAD(+)] (EC 1.1.1.8) C5L36_0E02100 UP000249293: Chromosome 5 415
| GO Term | Evidence | Action | Reason |
|---|---|---|---|
| GO:0005634 nucleus | IEA GO_REF:0000118 | KEEP AS NON CORE | Summary: Nucleus from TreeGrafter, descending from a fungi-scoped PAINT node whose only seed is S. cerevisiae GPD1, where the nuclear pool is a stress-induced relocalisation of an otherwise cytosolic enzyme. Reason: The transfer is phylogenetically in scope and is not rejected. PAINT places nucleus at PTHR11728 node PTN000919785 under a taxon:4751 (Fungi) constraint, and this Saccharomycotina protein descends from it, so the question is not whether the clade inherited the property but what the property is. The seeding observation is that Gpd1p, a cytosolic and peroxisomal enzyme, redistributes to the cytosol and nucleus when cells are osmotically stressed; the nuclear pool is conditional and is discussed as spatial regulation of redox potential rather than as a site of glycerol-3-phosphate synthesis. Nothing is known about the distribution of this protein in P. kudriavzevii, and the single seed is not a reason to doubt the node (a well-characterised seed can carry a node), only a reason to read what it actually shows. Kept, but as a conditional, secondary location: cytosol is where the core activity acts, and GO:0005634 is not listed among the core function's locations. Propagation Review Root cause: NO FAILURE NON CORE Failure modes: CONTEXT OR TISSUE MISMATCH Sources checked: PANTHER:PTN000919789 SOURCE WEAK OR INFERRED The TreeGrafter graft node for this protein. Its nucleus assertion descends from the PAINT IBD at PTN000919785, which is scoped to Fungi and seeded by SGD:S000002180 (S. cerevisiae GPD1). The node placement is sound for this target, but the underlying observation is a stress-conditional redistribution, so the transferred location is contextual rather than constitutive. Supporting Evidence: file:interpro/panther/PTHR11728/PTHR11728-paint.tsv PTHR11728 PTN000919785 GO:0005634 C IBD false SGD:S000002180 PMID:20026609 Exposure of cells to osmotic stress induces changes in the subcellular distribution of Gpd1p to the cytosol and nucleus. |
| GO:0005829 cytosol | IEA GO_REF:0000118 | ACCEPT | Summary: Cytosol from TreeGrafter; the NAD(+)-dependent enzyme of this family is cytosolic. Reason: PAINT asserts cytosol at PTHR11728 node PTN000198006 from experimental data in plants, flies, mammals and yeast, and the S. cerevisiae enzyme that defines the glycerol-synthesis role was characterised as cytosolic when it was cloned. The reviewed 388-residue model has no signal peptide, no transmembrane segment and no cleavable presequence in its record; that sub-argument is scoped to this model, because the reference-proteome model of the same locus carries a 27-residue N-terminal extension whose composition is presequence-shaped (see the alternative_records entry and suggested_questions). Independently of the N-terminus, the substrate dihydroxyacetone phosphate is a glycolytic intermediate, so a cytosolic site is where the reaction makes sense. This is the location of the core activity. Supporting Evidence: file:interpro/panther/PTHR11728/PTHR11728-paint.tsv PTHR11728 PTN000198006 GO:0005829 C IBD false PMID:8196651 We have cloned a gene encoding the key enzyme of glycerol synthesis, the NADH-dependent cytosolic glycerol-3-phosphate dehydrogenase, and we named it GPD1. |
| GO:0005975 carbohydrate metabolic process | IEA GO_REF:0000002 | MODIFY | Summary: Root-level carbohydrate metabolic process from the InterPro2GO mapping of IPR006109; true, but the specific polyol branch under it is what the evidence supports. Reason: Glycerol biosynthesis sits under carbohydrate metabolic process in GO (GO:0006114 is_a GO:0019401 is_a ... is_a GO:0005975), so the term is not wrong, but it is far less specific than the evidence allows. This enzyme catalyses the committed step of glycerol synthesis: it reduces the glycolytic intermediate dihydroxyacetone phosphate to glycerol 3-phosphate, which a phosphatase then converts to glycerol. The participation is direct, not substrate-like, and it is shown for this protein by complementation of S. cerevisiae gpd1 and gpd1 gpd2 mutants with glycerol accumulation, and in this species by the loss of glycerol when the gene is deleted. Refined to GO:0006114 glycerol biosynthetic process, which the gene does not yet carry. Proposed replacements: glycerol biosynthetic process Supporting Evidence: PMID:18452539 Functional analysis revealed that Saccharomyces cerevisiae gpd1Delta and gpd1Delta/gpd2Delta osmosensitive mutants transformed with CgGPD were restored to the wild-type phenotype when cultured in high osmolarity media, suggesting that it is a functional GPD protein. PMID:18452539 Transformants also accumulated glycerol intracellularly and GPD-specific activity increased significantly when stressed with NaCl PMID:41617712 Since glycerol-3-phosphate dehydrogenase (GPD) plays a key role in the glycerol biosynthesis pathway, we knocked out the GPD gene in strain JL01, generating the strain JL02. PMID:41617712 without any detectable glycerol accumulation |
| GO:0006072 glycerol-3-phosphate metabolic process | IEA GO_REF:0000120 | MODIFY | Summary: Direction-neutral glycerol-3-phosphate metabolic process; the evidence specifies which direction this enzyme runs. Reason: Correct but less specific than the evidence supports. Glycerol 3-phosphate is this enzyme's product, not its substrate, in vivo: in yeast the NAD(+)-dependent enzyme reduces dihydroxyacetone phosphate, while the reverse oxidation is carried by the FAD-dependent mitochondrial enzyme. Complementation of gpd1 mutants with glycerol accumulation (this protein) and the loss of glycerol on gene deletion (this species) both place it on the synthetic side. Refined to the is_a descendant GO:0046167 glycerol-3-phosphate biosynthetic process, the same replacement proposed for the catabolic row below. Proposed replacements: glycerol-3-phosphate biosynthetic process Supporting Evidence: PMID:18452539 Functional analysis revealed that Saccharomyces cerevisiae gpd1Delta and gpd1Delta/gpd2Delta osmosensitive mutants transformed with CgGPD were restored to the wild-type phenotype when cultured in high osmolarity media, suggesting that it is a functional GPD protein. PMID:18452539 Transformants also accumulated glycerol intracellularly and GPD-specific activity increased significantly when stressed with NaCl PMID:41617712 Since glycerol-3-phosphate dehydrogenase (GPD) plays a key role in the glycerol biosynthesis pathway, we knocked out the GPD gene in strain JL01, generating the strain JL02. |
| GO:0016616 oxidoreductase activity, acting on the CH-OH group of donors, NAD or NADP as acceptor | IEA GO_REF:0000002 | MODIFY | Summary: Grandparent oxidoreductase term from the InterPro2GO mapping of IPR011128, the NAD-dependent G3PDH N-terminal domain. Reason: True and correctly derived, but two levels above what the record supports. GO:0141152 glycerol-3-phosphate dehydrogenase (NAD+) activity, which this gene already carries and which this review accepts, is a descendant of GO:0047952, itself a descendant of this term. The substrate and cofactor are both fixed by the record: UniProt annotates the Rhea reaction with NAD(+) and sn-glycerol 3-phosphate, and the protein matches the NCBIfam model for NAD-dependent glycerol-3-phosphate dehydrogenase specifically. Refined to the specific term. Proposed replacements: glycerol-3-phosphate dehydrogenase (NAD+) activity Supporting Evidence: file:PICKU/GPD/GPD-uniprot.txt Reaction=sn-glycerol 3-phosphate + NAD(+) = dihydroxyacetone phosphate file:PICKU/GPD/GPD-uniprot.txt NCBIfam; TIGR03376; glycerol3P_DH; 1. |
| GO:0042803 protein homodimerization activity | IEA GO_REF:0000002 | KEEP AS NON CORE | Summary: Homodimerization transferred by InterPro2GO from IPR017751, the eukaryotic NAD-dependent glycerol-3-phosphate dehydrogenase signature. Reason: Defensible at the family level and kept, but not accepted as core. The protein matches IPR017751, the eukaryote-specific signature of this family, and the characterised eukaryotic member whose structure is known, human GPD1, was crystallised as a dimer (PMID:16460752; UniProt records the homodimer on P21695 from that work, though the paper's abstract does not state it). No oligomeric state has been measured for this protein or for any P. kudriavzevii member, so the claim rests on family membership alone. Self-association is in any case a structural property of the enzyme rather than a function it performs, so it is not part of the core function even if confirmed. Supporting Evidence: file:PICKU/GPD/GPD-uniprot.txt InterPro; IPR017751; G3P_DH_NAD-dep_euk. file:PICKU/GPD/GPD-uniprot.txt PIRSF; PIRSF000114; Glycerol-3-P_dh; 1. |
| GO:0046168 glycerol-3-phosphate catabolic process | IEA GO_REF:0000120 | MODIFY | Summary: Catabolic direction asserted by a UniProt rule; the enzyme runs the opposite way in vivo, synthesising glycerol 3-phosphate rather than breaking it down. Reason: This is a direction error, not a granularity problem, so the replacement is a sibling rather than a descendant. The rule appears to read the direction off the EC 1.1.1.8 reaction as written (sn-glycerol 3-phosphate + NAD+ = dihydroxyacetone phosphate + NADH), but the physiological flux through the NAD(+)-dependent enzyme in yeast is reductive. S. cerevisiae cells lacking both NAD(+)-dependent isoenzymes make no detectable glycerol, which is only possible if these enzymes produce glycerol 3-phosphate rather than consume it; the enzyme that oxidises glycerol 3-phosphate in yeast is the separate FAD-dependent mitochondrial protein. For this protein specifically, complementation of gpd1 mutants restores glycerol accumulation, and deleting the gene in P. kudriavzevii removes glycerol from the fermentation. Replaced by GO:0046167 glycerol-3-phosphate biosynthetic process. The same correction is recorded as a source defect, since the rule will keep propagating the catabolic term to every member of this family. Proposed replacements: glycerol-3-phosphate biosynthetic process Supporting Evidence: PMID:9171333 Mutants deleted for both GPD1 and GPD2 do not produce detectable glycerol PMID:8196651 We have cloned a gene encoding the key enzyme of glycerol synthesis, the NADH-dependent cytosolic glycerol-3-phosphate dehydrogenase, and we named it GPD1. PMID:18452539 Transformants also accumulated glycerol intracellularly and GPD-specific activity increased significantly when stressed with NaCl PMID:41617712 without any detectable glycerol accumulation |
| GO:0047952 glycerol-3-phosphate dehydrogenase [NAD(P)+] activity | IEA GO_REF:0000118 | MODIFY | Summary: The cofactor-agnostic NAD(P)+ form of the activity, from the PAINT node for this family. Reason: Correct but less specific than the record supports. GO:0047952 covers enzymes that use NAD+ or NADP+; this protein's cofactor is fixed as NAD(+) by UniProt's Rhea-linked catalytic activity and by the PIRSR-annotated NAD(+) binding residues, and it matches the NCBIfam model for the NAD-dependent subtype. Refined to its is_a descendant GO:0141152, which this gene already carries and which this review accepts as the core activity. This is a refinement, not a rejection: the PAINT node that asserts GO:0047952 is seeded by characterised members in yeast, fission yeast, fly, mouse and human, and is sound. Proposed replacements: glycerol-3-phosphate dehydrogenase (NAD+) activity Supporting Evidence: file:PICKU/GPD/GPD-uniprot.txt Reaction=sn-glycerol 3-phosphate + NAD(+) = dihydroxyacetone phosphate file:PICKU/GPD/GPD-uniprot.txt NCBIfam; TIGR03376; glycerol3P_DH; 1. file:PICKU/GPD/GPD-uniprot.txt /ligand="NAD(+)" |
| GO:0051287 NAD binding | IEA GO_REF:0000120 | KEEP AS NON CORE | Summary: NAD(+) binding by the N-terminal Rossmann-fold domain. Reason: Well supported and kept. UniProt annotates five NAD(+) binding positions by the curated PIRSR rule PIRSR000114-3, in and around the Pfam PF01210 Rossmann-fold domain, and the cloning paper reports the same domain organisation for this sequence. NAD(+) is the cosubstrate of the core reaction, so the binding term describes part of the catalytic mechanism rather than an independent function; it is kept as non-core beside GO:0141152. Supporting Evidence: file:PICKU/GPD/GPD-uniprot.txt /ligand="NAD(+)" file:PICKU/GPD/GPD-uniprot.txt /note="Glycerol-3-phosphate dehydrogenase NAD-dependent N- PMID:18452539 The CgGPD gene consisted of an N-terminal NAD(+)-binding domain and a central catalytic domain |
| GO:0141152 glycerol-3-phosphate dehydrogenase (NAD+) activity | IEA GO_REF:0000120 | ACCEPT | Summary: The core activity: NAD(+)-linked interconversion of dihydroxyacetone phosphate and sn-glycerol 3-phosphate, run reductively in vivo. Reason: Every line of evidence converges here. The record matches the NAD-dependent glycerol-3-phosphate dehydrogenase family by NCBIfam TIGR03376, PIRSF000114, PROSITE PS00957 and PANTHER PTHR11728, with both family domains intact and the PIRSR proton-acceptor lysine and NAD(+) and substrate binding residues present. Beyond family membership, there is direct functional evidence for this very sequence: expressed in S. cerevisiae gpd1 and gpd1 gpd2 mutants it restored growth under high osmolarity, the transformants accumulated glycerol, and GPD-specific enzyme activity rose under NaCl stress. Accepted as the core molecular function. The term is cofactor-specific and matches the annotated Rhea reaction, so no refinement applies. Supporting Evidence: file:PICKU/GPD/GPD-uniprot.txt Reaction=sn-glycerol 3-phosphate + NAD(+) = dihydroxyacetone phosphate file:PICKU/GPD/GPD-uniprot.txt NCBIfam; TIGR03376; glycerol3P_DH; 1. file:PICKU/GPD/GPD-uniprot.txt /ligand="NAD(+)" PMID:18452539 Functional analysis revealed that Saccharomyces cerevisiae gpd1Delta and gpd1Delta/gpd2Delta osmosensitive mutants transformed with CgGPD were restored to the wild-type phenotype when cultured in high osmolarity media, suggesting that it is a functional GPD protein. PMID:18452539 Transformants also accumulated glycerol intracellularly and GPD-specific activity increased significantly when stressed with NaCl |
Loading supporting content…
Download this section (compressed HTML)Q: Which initiator codon does this gene use in vivo? The reference-proteome model A0A2U9RB90 adds 27 N-terminal residues to the 388-residue product cloned and shown to be functional, and the extension is basic, Ser/Thr-rich and free of acidic residues, the composition of a mitochondrial targeting presequence, so it would be worth testing for mitochondrial import rather than assuming it is a gene-model artefact.
Q: Does P. kudriavzevii use a single NAD(+)-dependent glycerol-3-phosphate dehydrogenase for both the osmotic and the redox role that S. cerevisiae splits between GPD1 and GPD2? Only one family member is present in the reference proteome, and the regulation of that single gene has not been described.
Experiment: Purify the recombinant enzyme and measure steady-state kinetics in both directions with NAD(+)/NADH, to establish the catalytic preference directly rather than inferring it from genetics, and test whether NADP(H) is used at all (which would decide between GO:0141152 and the broader GO:0047952).
Experiment: Determine the subcellular distribution of a tagged GPD in P. kudriavzevii under normal and hyperosmotic conditions, to test whether the stress-induced nuclear redistribution seen for S. cerevisiae Gpd1p, which is the basis of the transferred nucleus annotation, occurs in this species. Score the peroxisomal pool in the same experiment: Gpd1p is imported into peroxisomes through an N-terminal PTS2 read by Pex7p, so co-localisation with a peroxisomal marker, and the behaviour of the protein in a pex7 background, would say whether that route is also used here. The N-terminus of A8W1U0 lies in the region where Gpd1p's PTS2 sits, and the 27-residue extension of the reference-proteome model falls in the same window, so the two gene models should be tested separately.
Experiment: Compare the two gene models by 5' RACE or ribosome profiling, and express both the 388- and 415-residue forms in a gpd-null background, to establish which product is made and whether the N-terminal extension changes localisation or activity.
Loading supporting content…
Download this section (compressed HTML)Loading supporting content…
Download this section (compressed HTML)