FMO3

Summary: The existing experimental dimethylaniline assertion is retained with source-specific access uncertainty.

Reason: Retain the existing curated N,N-dimethylaniline reaction as a contextual FMO activity. The actual GO definition is substrate-specific; TMA oxidation is not itself direct evidence for this reaction. Reviewed UniProt maps RHEA:24468 and PAINT places a positive ancestral assertion on the FMO3 path. The PMID:9536088 substrate-specific experiment was not independently inspected, and the UniProt/GO mappings are not separate experiments. Do not infer a false annotation or ancestral loss from this access limit; the defining core is supported TMA oxygenation.

Summary: The verified ancestral placement supports contextual N,N-dimethylaniline activity.

Reason: Retain the existing curated N,N-dimethylaniline reaction as a contextual FMO activity. The actual GO definition is substrate-specific; TMA oxidation is not itself direct evidence for this reaction. Reviewed UniProt maps RHEA:24468 and PAINT places a positive ancestral assertion on the FMO3 path. The PMID:9536088 substrate-specific experiment was not independently inspected, and the UniProt/GO mappings are not separate experiments. Do not infer a false annotation or ancestral loss from this access limit; the defining core is supported TMA oxygenation.

Summary: The specific dimethylaniline reaction is retained outside the principal TMA core.

Reason: Retain the existing curated N,N-dimethylaniline reaction as a contextual FMO activity. The actual GO definition is substrate-specific; TMA oxidation is not itself direct evidence for this reaction. Reviewed UniProt maps RHEA:24468 and PAINT places a positive ancestral assertion on the FMO3 path. The PMID:9536088 substrate-specific experiment was not independently inspected, and the UniProt/GO mappings are not separate experiments. Do not infer a false annotation or ancestral loss from this access limit; the defining core is supported TMA oxygenation.

Summary: Human FMO3–CREB3 interaction supports DNA-binding transcription factor binding.

Reason: PMID:25910212 has three exact positive IntAct records for human P31513 and O43889-2, covering pooled screening, pairwise yeast two-hybrid testing and the validated-two-hybrid record. Selected actual Results and Methods describe human ORF interaction screening and allele comparisons; the target records distinguish FMO3 substitutions that retain or disrupt the interaction. Reviewed UniProt identifies O43889-2 as the displayed CREB3/LZIP product named isoform 1, a CREB-family DNA-binding transcription-factor precursor. GO:0140297 therefore describes the supported partner class more informatively than protein binding. A narrower CREB-protein binding term remains a candidate rather than the selected replacement because its CREB3 family boundary is not resolved by the bounded usage check. This does not establish FMO3 transcription-factor activity, binding to a DNA-bound or processed CREB3 form, a transcription-regulatory mechanism, or endogenous mammalian co-localization. The paper-level orthogonal GPCA validation is not claimed for this particular pair; the original supplement cells and insert sequences were not independently reconstructed.

Proposed replacements: DNA-binding transcription factor binding

Summary: The source-specific FMO3–CREB3 isoform pair supports DNA-binding transcription factor binding.

Reason: PMID:31515488 has an exact positive IntAct yeast two-hybrid record for P31513 and O43889-2; its CREB3 variant features record interaction perturbation. The selected Methods specify pairwise wild-type/mutant human ORF assays, autoactivation controls and reproducibility criteria. A second FMO3 record uses unsuffixed O43889 and is retained separately rather than silently equated with the requested isoform. Reviewed UniProt establishes that O43889-2 is the displayed CREB3/LZIP product named isoform 1, supporting GO:0140297 as the partner-class refinement. GO:0008140 is retained as an unselected narrower candidate pending resolution of its CREB3 family scope. These data do not establish FMO3-mediated transcription regulation, a processed-CREB3-specific interaction, mammalian endogenous binding or a target-specific orthogonal PCA assay. Original supplementary-table cells and complete construct sequences were not independently audited.

Proposed replacements: DNA-binding transcription factor binding

Summary: Binary association with CLEC10A supports signaling receptor binding.

Reason: Three positive IntAct records from PMID:32296183 support human FMO3–CLEC10A interaction capacity in yeast ORF-fusion assays, using prey pooling, array testing and validated two-hybrid records. The source describes pairwise retesting and sequence confirmation; these are records from one publication, not independent studies. Reviewed CLEC10A Q8IUN9 is a signaling receptor, so GO:0005102 supplies the supported partner class. The official term comment treats receptor agonist activity as a separate annotation to consider where appropriate; receptor binding alone does not assert agonism. The ER/hepatic FMO3 and myeloid cell-surface/endolysosomal CLEC10A contexts leave native encounter and physiological relevance unestablished. No endogenous co-expression, co-localization, signaling activation, glycan-ligand role or pair-specific MAPPIT/GPCA validation is claimed, and the distinct contexts do not by themselves disprove the measured interaction capacity.

Proposed replacements: signaling receptor binding

Summary: Binary association with SGSM3 supports GTPase activating protein binding.

Reason: PMID:32296183 has three exact positive IntAct records for human FMO3 and SGSM3 Q96HU1 in the yeast two-hybrid screen, array retest and validated-two-hybrid dataset. Reviewed SGSM3 function identifies a Rab5 GTPase-activating protein, rather than a small GTPase. GO:0032794 therefore captures the established partner class. This refinement does not assign GAP activity, small-GTPase binding, Rab regulation or an endocytic process to FMO3. The reported human ORF fusion assays and pairwise screening controls support the interaction capacity; endogenous physiological association, complete insert sequence and pair-specific orthogonal validation were not established by this consultation.

Proposed replacements: GTPase activating protein binding

Summary: Binary association with SACM1L supports phosphatase binding.

Reason: PMID:32296183 has three exact positive IntAct records for human FMO3 and SACM1L Q9NTJ5, including prey-pooling, array and validated yeast two-hybrid records. Reviewed SACM1L annotation identifies a phosphoinositide phosphatase, supporting GO:0019902. Protein phosphatase binding would incorrectly narrow the partner substrate class, and the interaction does not give FMO3 phosphatase activity or establish regulation of phosphoinositide metabolism. The assays use human ORF fusion constructs in yeast; native mammalian interaction, complete insert reconstruction and pair-specific orthogonal validation remain outside the inspected evidence.

Proposed replacements: phosphatase binding

Summary: The electronic mapping agrees with the reviewed ER-membrane localization.

Reason: Retain the ER-membrane location. The electronic SL-0097 mapping agrees with the reviewed human subcellular-location record and verified rabbit FMO3 source lineage, as well as the explicit human Reactome catalyst entity. This is a curated localization mapping, not another primary localization experiment; do not infer a membrane-side or topology claim.

Summary: Verified rabbit FMO3 transfer supports the conserved ER-membrane location.

Reason: Retain the ER-membrane localization as core. The original ISS donor P32417 is verified rabbit FMO3. Reviewed human P31513 explicitly transfers microsome and ER-membrane localization from that record; the donor record traces microsomal protein purification/sequencing to its source literature. This supports the conserved location, not an independently observed human imaging experiment, cytosolic-versus-luminal catalytic orientation, or measured single-pass topology.

Summary: The curated human Reactome catalyst explicitly occupies the ER membrane.

Reason: Retain the location: the original Reactome event R-HSA-139970 leads through catalyst114712 to FMO3:FAD complex217276 and human P31513 protein54929, explicitly located in GO:0005789. Event substrates also occupy the ER lumen, but their compartment is not substituted for the enzyme location. The source is a curated human reaction model rather than an independently verified microscopy experiment.

Summary: The disease reaction preserves the FMO3 ER-membrane compartment.

Reason: Retain the ER-membrane location: the disease event R-HSA-5602966 is a FailedReaction whose catalyst links to the human FMO3-mutant/FAD complex9631787 and set5603019. All five member proteins refer to human P31513 and GO:0005789. The failed-reaction title concerns deficient TMA catalysis, not loss of enzyme localization and not a NOT localization assertion. This does not imply every allele has normal abundance or activity.

Summary: Independent human FMO3 product assays corroborate the original TMA annotation.

Reason: FMO3 directly oxygenates trimethylamine to TMAO. Recombinant human protein and human adult-liver microsomes support this substrate-specific activity; independent membrane-fraction assays of human FMO3 further support product formation. The disease-associated variants are not separate normal products. PMID:9536088 specific assay details remain uninspected; acceptance is also supported independently by PMID:9776311 and PMID:17531949.

Summary: Recombinant human FMO3 and adult-liver microsomes oxidize trimethylamine.

Reason: FMO3 directly oxygenates trimethylamine to TMAO. Recombinant human protein and human adult-liver microsomes support this substrate-specific activity; independent membrane-fraction assays of human FMO3 further support product formation. The disease-associated variants are not separate normal products. PMID:9536088 specific assay details remain uninspected; acceptance is also supported independently by PMID:9776311 and PMID:17531949.

Summary: FMO3 catalyzes NADPH-dependent conversion of trimethylamine to TMAO.

Reason: FMO3 directly oxygenates trimethylamine to TMAO. Recombinant human protein and human adult-liver microsomes support this substrate-specific activity; independent membrane-fraction assays of human FMO3 further support product formation. The disease-associated variants are not separate normal products. PMID:9536088 specific assay details remain uninspected; acceptance is also supported independently by PMID:9776311 and PMID:17531949.

Summary: The on-path ancestral taurine-synthesis assertion is biologically supported and non-core.

Reason: The full primary study directly tested human FMO3 baculosomes and detected conversion of hypotaurine to taurine, at much lower activity than FMO1. This is real catalytic participation despite the FMO1-focused title. Retain as non-core; the mouse knockout/major physiological argument and reported FMO1 kinetic constants are not FMO3-specific evidence. For the IBA, the positive on-path PAINT placement is retained without asserting loss or treating the target donor as circular.

Summary: FMO3 directly contributes a chemical step to taurine synthesis in the tested assay.

Reason: The full primary study directly tested human FMO3 baculosomes and detected conversion of hypotaurine to taurine, at much lower activity than FMO1. This is real catalytic participation despite the FMO1-focused title. Retain as non-core; the mouse knockout/major physiological argument and reported FMO1 kinetic constants are not FMO3-specific evidence.

Summary: Human recombinant FMO3 generates albendazole sulfoxide.

Reason: Human recombinant FMO3 microsomes generate albendazole sulfoxide above control, supporting this specific drug-metabolizing activity. Human-liver studies indicate a larger CYP contribution, principally CYP3A4, and do not make FMO3 the exclusive route. Retain as a contextual xenobiotic substrate activity; TMA oxygenation defines the principal core.

Summary: Albendazole S-oxygenation is a demonstrated contextual drug-metabolizing activity.

Reason: Human recombinant FMO3 microsomes generate albendazole sulfoxide above control, supporting this specific drug-metabolizing activity. Human-liver studies indicate a larger CYP contribution, principally CYP3A4, and do not make FMO3 the exclusive route. Retain as a contextual xenobiotic substrate activity; TMA oxygenation defines the principal core.

Summary: Human FMO3 baculosomes produce taurine from hypotaurine.

Reason: The full primary study directly tested human FMO3 baculosomes and detected conversion of hypotaurine to taurine, at much lower activity than FMO1. This is real catalytic participation despite the FMO1-focused title. Retain as non-core; the mouse knockout/major physiological argument and reported FMO1 kinetic constants are not FMO3-specific evidence.

Summary: Hypotaurine oxidation is a demonstrated secondary substrate activity.

Reason: The full primary study directly tested human FMO3 baculosomes and detected conversion of hypotaurine to taurine, at much lower activity than FMO1. This is real catalytic participation despite the FMO1-focused title. Retain as non-core; the mouse knockout/major physiological argument and reported FMO1 kinetic constants are not FMO3-specific evidence.

Summary: FAD is the bound catalytic cofactor of FMO3.

Reason: FAD is the catalytic prosthetic cofactor of FMO3. Direct biochemical analysis of human FMO3 measures FAD incorporation/retention, and purified protein spectroscopy follows reduction and oxygenation of the bound flavin. This is specific cofactor binding supporting the enzyme mechanism, not generic protein binding.

Summary: Human FMO3 directly binds the NADP cofactor.

Reason: Purified human FMO3 directly binds NADP+, measured by calorimetry, while NADPH supplies reducing equivalents to FAD. The retained oxidized cofactor stabilizes the oxygenating intermediate; normal protein is distinguished from the N61S disease variant.