The fog (flavoglaucin) gene cluster contains 9 genes (fogA-fogI) and was characterized by Nies et al. (2020) [PMID:32134669, DOI:10.1021/acs.orglett.0c00440, "The biosynthetic pathway of the prenylated salicylaldehyde flavoglaucin and congeners in Aspergillus ruber was elucidated by genome mining, heterologous expression, precursor feeding, and biochemical characterization"]. The cluster produces prenylated salicylaldehyde derivatives: flavoglaucin, dihydroauroglaucin, auroglaucin, and aspergin.
The genome of A. ruber CBS 135680 was sequenced by Kis-Papo et al. (2014) [PMID:24811710, DOI:10.1038/ncomms4745, "Genomic adaptations of the halophilic Dead Sea filamentous fungus Eurotium rubrum"].
According to the UniProt entry and PMID:32134669:
Key insight from Nies et al.: "The polyketide skeleton was released as alkylated salicyl alcohols, which is a prerequisite for consecutive hydroxylation and prenylation, before reoxidation to the final aldehyde products" PMID:32134669.
"Impairs the production of flavoglaucin and congeners, and leads to very low accumulation of the carboxylic acid (8E,10E,12E)-3,5,7-trihydroxytetradeca-8,10,12-trienoic acid" [PMID:32134669, via UniProt annotation].
fogD is assigned to PANTHER family PTHR44169 (NADPH-dependent 1-acyldihydroxyacetone phosphate reductase), subfamily PTHR44169:SF3 (short-chain dehydrogenase SrdE). TreeGrafter propagates GO annotations from well-characterized members of this family, which are lipid metabolism enzymes (Ayr1p-like DHAP reductases in yeast).
This is a clear case of incorrect annotation transfer across a functionally diverse enzyme superfamily. The SDR superfamily is enormous and catalyzes many unrelated reactions. fogD's actual function is in secondary metabolite (polyketide) biosynthesis, not lipid metabolism.
All six TreeGrafter annotations derive from PANTHER:PTN001211783 via GO_REF:0000118:
GO:0000140 (acylglycerone-phosphate reductase (NADP+) activity) -- WRONG. This is the activity of Ayr1p, a lipid biosynthesis enzyme. fogD has no demonstrated or predicted DHAP reductase activity.
GO:0004806 (triacylglycerol lipase activity) -- WRONG. This is a hydrolase activity; fogD is an oxidoreductase. Completely wrong reaction type.
GO:0005783 (endoplasmic reticulum) -- WRONG. No evidence for ER localization. Secondary metabolite cluster enzymes in fungi typically function in the cytoplasm.
GO:0005811 (lipid droplet) -- WRONG. No evidence for lipid droplet localization. Propagated from lipid metabolism family members.
GO:0006654 (phosphatidic acid biosynthetic process) -- WRONG. fogD has no role in phospholipid biosynthesis.
GO:0019433 (triglyceride catabolic process) -- WRONG. fogD has no role in triglyceride catabolism.
The six TreeGrafter rejections above are superseded by UNDECIDED pending a focused OpenScientist investigation. Secondary-metabolite participation does not prove exclusive function. Ayr1 is a direct counterexample to the original claim that no SDR can be a triglyceride lipase (PMID:24187129(https://pubmed.ncbi.nlm.nih.gov/24187129/)); its primary study establishes source activity, not FogD activity. Lack of transmembrane or secretion features cannot exclude peripheral ER or lipid-droplet association.
The current PAINT slice records experimentally grounded Ayr1-seeded IBDs, but does not contain historical node PTN001211783. The ER assertion is at PTN001213826; the other five claims are at PTN001963740, each seeded by SGD:S000001386. Historical graft reconstruction is unresolved. A single well-characterized source is valid ancestral evidence, not intrinsically weak support.
The main PMID:32134669 cache is abstract-only and the publisher main article returned 403. Its open supplement was independently downloaded from the ACS Figshare API (article 11944365, file 21927279; DOI:10.1021/acs.orglett.0c00440.s001; MD5 a4e37dcbc2f110a230a24b385e1cfba6). Methods S5-S6 identify A. ruber QEN-0407-G2 as the DNA/RNA source and CBS135680 as the reference genome. Table S1 maps FogD to EYE95338/EURHEDRAFT_455854. The supplement identifies JN009 as the fogD deletion strain in the reconstructed A. nidulans JN004 cluster background. Table S2 and Figure S8 (S18/S30) were visually inspected: the latter contains a small 2d peak in the deletion strain. This corrects any implication of a native CBS135680 deletion or purified FogD catalytic assay.
Retain broad secondary-metabolite participation and a conservative SDR-derived oxidoreductase proposal. The specific CH-OH/NAD(P) reaction assignment was withdrawn pending direct reaction/cofactor evidence; this caution was superseded on 2026-09-27 by provisional adoption of GO:0016616 from UniProt's EC=1.1.1.- placement. Remove the redundant NEW polyketide-process row because it descends from the existing secondary-metabolite process; keep the specific pathway in the core summary. The former flavoglaucin-term proposal is moved to an ontology question. Unverified supporting_text_fulltext paraphrases are removed, and UniProt-derived statements are explicitly attributed to UniProt rather than invented primary quotations.
Assessed fogD-hypotheses/sdr-lipid-functions-and-polyketide-release/openscientist.md.
The report validated the concern that GO:0000140, GO:0004806, GO:0005783,
GO:0005811, GO:0006654 and GO:0019433 are Ayr1-seeded TreeGrafter transfers
without FogD-specific evidence. Updated the acyl-DHAP reductase row to MODIFY
toward GO:0016616, because UniProt's EC=1.1.1.- placement supports CH-OH
donor/NAD(P) acceptor oxidoreductase chemistry but not the Ayr1-specific
acyl-DHAP substrate. The other Ayr1 lipid activity, location and process rows
remain non-core or over-annotated rather than reverting to the earlier REMOVE
decisions because the source Ayr1 annotations are experimentally supported.
retired: true; its ACCEPT