Focused Curation Report — Pyricularia oryzae PoMZ_10221 / F8U970

Hypothesis slug: rhamnose-enzyme-and-mat-regulatory-capacity
Focus type: function_assignment
Gene: PoMZ_10221 (UniProt F8U970, Swiss-Prot reviewed; EMBL JF740057 / AEH41994.1; genome QBZ54521.1)
Organism: Pyricularia oryzae (= Magnaporthe grisea/oryzae), NCBITaxon:318829

Executive Judgment

Verdict on the seed hypothesis: REFUTED (for the MAT claims) / the UDP-rhamnose claim is SUPPORTED.

The seed proposes that F8U970 retains methionine adenosyltransferase (MAT) regulator activity, MAT-complex membership, and participation in S-adenosylmethionine (SAM) biosynthesis in addition to its experimentally supported UDP-rhamnose epimerase/reductase activity, and asks that each claim be adjudicated independently.

• The UDP-rhamnose epimerase/reductase function is directly and experimentally supported (IDA, PMID 22102281). This is the protein's primary, catalytically demonstrated function.
• The three MAT-related claims (MAT regulator activity, MAT complex, SAM biosynthesis) are not supported by any direct, interaction, localization, or mutant evidence. On F8U970 they exist only as IEA:TreeGrafter automated annotations propagated through PANTHER PTHR10491:SF4 ("Methionine adenosyltransferase 2 subunit beta"). They are best explained as phylogenetic-propagation (paralog-lumping) over-annotations, not a genuine second capacity.

Most important caveat: "Not proven" is not the same as "proven absent." No experiment has tested MAT-binding by F8U970. However, the burden of evidence, the twilight-zone homology to MAT2B (22%), the strong homology to true UDP-rhamnose synthases (63%), the absence of any MAT/SAM mention in the primary paper, and the biological implausibility of a MAT2B-type beta regulatory subunit in fungi together make the "retained MAT capacity" claim unsupported for curation purposes.

The seed is also correct on one negative point: this protein is not the fungal dihydroorotate dehydrogenase claimed by an earlier automated report — UniProt and the primary literature unambiguously assign UDP-rhamnose synthesis.

Evidence Matrix

GO Curation Implications

Lead (requires curator verification):

• RETAIN (experimental, core):
• MF GO:0010489 UDP-4-keto-6-deoxy-glucose-3,5-epimerase activity (IDA)
• MF GO:0010490 UDP-4-keto-rhamnose-4-keto-reductase activity (IDA)
• BP GO:0019300 rhamnose biosynthetic process (IDA)

• (Consider also a general nucleotide-sugar biosynthetic process / dTDP-4-dehydrorhamnose reductase family MF if the curation model prefers it; NADP binding CC/MF is defensible from the reductase reaction.)

• REMOVE / DO NOT PROPAGATE (over-annotation; IEA:TreeGrafter only):

• CC GO:0048269 methionine adenosyltransferase complex
• MF GO:0048270 methionine adenosyltransferase regulator activity
• BP GO:0006556 S-adenosylmethionine biosynthetic process

Rationale: no experimental/interaction/localization support; artifact of PTHR10491:SF4 lumping the fungal UDP-rhamnose enzyme with the divergent metazoan MAT2B regulator; biologically implausible in a fungus that lacks a MAT2B-type beta regulatory subunit. Crucially, SAM biosynthesis in P. oryzae is performed by a separate dedicated canonical SAM synthase (G4NCL5, EC 2.5.1.6, PTHR11964), so GO:0006556 on F8U970 is redundant and incorrect. A co-family paralog (G4ML22) carries the same spurious SF4 "MAT2 beta" tag, confirming the mislabel is systematic.

• Do not recommend "protein binding" — a specific, supported MF (the epimerase/reductase activities above) is available.

The evidence supports MF + BP (catalysis and rhamnose biosynthesis) as core, and argues the MAT MF/BP/CC as non-core and removable/excludable (candidate NOT/exclusion or simply do-not-propagate).

Mechanistic Scope

Immediate molecular function (direct): F8U970 catalyzes the committed epimerization + NADPH-dependent reduction step converting UDP-4-keto-6-deoxy-D-glucose → UDP-β-L-rhamnose (RHEA:56888 then RHEA:56892), the second enzyme of a two-gene fungal UDP-rhamnose pathway.

Downstream/context (not the tested molecular activity): rhamnose-containing glycans influence host–pathogen interactions (adhesion, recognition, virulence, biofilm). These are pathway/phenotype consequences, not molecular functions of this protein, and their tissue-specific regulation is a transcriptional observation.

The MAT "function" would require physical association with a MAT catalytic subunit and modulation of SAM synthase activity — none of which is demonstrated. It is inferred only from shared fold via automated propagation.

Conflicts and Alternatives

• Paralog/family lumping (primary alternative explanation): PANTHER PTHR10491 contains both sugar-nucleotide epimerase-reductases (RmlD/RHM-type) and the metazoan MAT2B regulator, which share the extended-SDR / NAD(P)-binding Rossmann fold. TreeGrafter grafts F8U970 near a node carrying MAT2B-derived IBDs, exporting MAT terms onto a protein whose real neighbors are UDP-rhamnose synthases (63% identity to RHM1). The seed's own notes ("exact fungal accession absent from reference tree," "broad regulator IBD PTN000051877 not in the recovered graft path") are consistent with an unstable/over-reaching graft.
• Organism-specific difference: the source of the propagated IBDs, human MAT2B (Q9NZL9), is characterized specifically as the mammalian regulatory β-subunit of MATII (PMID 31496615: "Mammalian systems express two genes"), and its regulatory action operates through protein–protein interactions in metazoan signaling (binds GIT1, activates ERK; PMID 32046373). Fungal methionine adenosyltransferases (e.g., S. cerevisiae Sam1/Sam2) are single-subunit homo-oligomeric SAM synthetases without a MAT2B-type β regulatory subunit, so there is no α2β2 MAT complex for F8U970 to join in P. oryzae. (Stated as inference from the mammalian-restricted MAT2B literature; not directly disproven experimentally.)
• Database carry-over: a prior automated ("OpenAI") report mislabeled the protein as dihydroorotate dehydrogenase; that is a separate, clearly incorrect assignment and should not influence curation.
• No competing experimental evidence supports MAT function.

Knowledge Gaps

1. Direct test of MAT association. Checked: no interaction/localization data in UniProt or the primary paper. Matters because the seed asks specifically about "catalytic/regulatory coexistence." Resolved by: co-IP / pull-down of F8U970 with P. oryzae SAM synthetase, or AlphaFold-Multimer interface scoring — neither currently in evidence.
2. Exact TreeGrafter version/graft path. Checked: UniProt reports the SF4 assignment and IEA terms but not the frozen graft node internals; I could not programmatically resolve the specific PTN008946252 vs PTN000051875/PTN000051877 IBD propagation logic. Matters for reproducing exactly which IBD exported the MAT terms. Resolved by: inspecting the PAINT/TreeGrafter annotation for PTHR10491 at the frozen version.
3. NADP cofactor CC/MF specificity. The reductase uses NADPH (RHEA:56892) but an explicit NADP binding MF is not asserted; low-priority, easily added.

Discriminating Tests

1. AlphaFold-Multimer / co-IP: test whether F8U970 forms a complex with the P. oryzae methionine adenosyltransferase catalytic subunit. Predicted result if hypothesis false: no stable interface (unlike human MAT2A–MAT2B).
2. Reciprocal-best-hit / phylogeny with proper outgroups: place F8U970 among fungal/plant UDP-rhamnose synthases vs MAT2B clade; expect robust grouping with RHM-type enzymes (consistent with 63% identity), not MAT2B.
3. Δgene metabolite phenotype: knockout should reduce UDP-rhamnose / rhamnoglycans (already implied by pathway), and should not perturb SAM pools if MAT function is absent.
4. Active-site residue audit: confirm RmlD reductase catalytic/NADP-binding residues are intact (they are, given retained catalysis), whereas MAT2B-specific regulatory-interface residues are absent.

Curation Leads (verify before applying)

• Action change: downgrade/remove the MAT-related IEA:TreeGrafter terms (GO:0048269, GO:0048270, GO:0006556) as paralog-driven over-annotations; retain the IDA UDP-rhamnose terms as core.
• Candidate reference to cite for core function: PMID 22102281 — verify snippet: "The second gene encodes a bifunctional UDP-4-keto-6-deoxyglucose-3,5-epimerase/-4-reductase that converts UDP-4-keto-6-deoxyglucose to UDP-rhamnose."
• Candidate GO (retain/core): GO:0010489, GO:0010490 (MF); GO:0019300 (BP). Optionally add NADP binding.
• Candidate GO (exclude/non-core): GO:0048269, GO:0048270, GO:0006556.
• Suggested curator question: Is there any organism-specific evidence (interaction, complexome, localization) that a MAT2B-type beta regulator exists in P. oryzae? If not, the CC term is untenable.
• Suggested experiment: AlphaFold-Multimer/co-IP of F8U970 with P. oryzae SAM synthetase to formally close the "coexistence" question.
• Note: flag and discard the earlier "dihydroorotate dehydrogenase" automated assignment as incorrect.

Provenance

• Primary literature: PMID 22102281 abstract (PubMed).
• Records: UniProt Swiss-Prot F8U970 JSON (GO evidence codes, catalytic activities, PANTHER/InterPro/Pfam DR lines).
• Computed: global Needleman–Wunsch (BLOSUM62) pairwise identities — F8U970 vs Q9NZL9 (22.0%), P31153 (25.7%), P19358 (24.9%), Q9SYM5 (62.8%). Executed code and outputs are retained in this iteration's execution log.
• Computed (Iteration 3): UniProt proteome query of P. oryzae (taxon 318829) — the organism's canonical SAM synthase is G4NCL5 (EC 2.5.1.6, Pfam PF00438/PF02772/PF02773, PTHR11964), distinct from F8U970; RmlD-fold paralog G4ML22 (PF04321) shares the same spurious PTHR10491:SF4 "MAT2 beta" label. Executed code and outputs retained in the Iteration 3 execution log.

Decision Table (leads — verify before applying)

Artifacts

• OpenScientist final report
• OpenScientist final report