B8BAB0 (Oryza sativa subsp. indica) — Evaluation of ProtNLM2 "pollen maturation (GO:0010152)" prediction
Focus type: computational_prediction · Term under test: pollen maturation (GO:0010152) Gene: B8BAB0 / OsI_29064 / Gramene BGIOSGA028580 (chromosome 8) · UniProt PE: 4 (Predicted) · Curated GO annotations: none Japonica ortholog: OsBURP12 / UniProt Q6ZA27 (86.7% id) = RAP-DB Os08g0380100 (KEGG dosa:Os08g0380100) — use this locus for RiceXPro/RAP-DB expression lookup.
Executive Judgment
Verdict: REFUTED / over-annotated for the specific term GO:0010152 (pollen maturation).
The ProtNLM2 prediction of pollen maturation is not supported by domain architecture, subfamily placement, orthology, or structure. B8BAB0 is unambiguously a member of the PG1β-like (polygalacturonase non-catalytic β-subunit-like) subfamily of BURP-domain proteins — a secreted cell-wall / pectin-associated clade — and is not a member of the cereal anther-specific RAFTIN / BURP-VII clade that supplies the only well-documented "reproductive development" precedent among BURP proteins. The prediction appears to be a frequency/superfamily over-generalization: it borrows the "some BURP proteins act in reproduction" association (true only for the distinct RAFTIN clade) and applies it to a protein that actually belongs to the cell-wall clade.
Most important caveat: I could not retrieve B8BAB0-specific pollen/anther expression data programmatically. Sequence and structure cannot, by themselves, exclude tissue-specific expression; however, the seed hypothesis asks whether a specific pollen-maturation process can be established from sequence — and it cannot. If anything, the supportable homology leads point to extracellular/cell-wall localization and pectin/cell-wall metabolism, not pollen maturation.
Evidence Matrix
| # | Citation | Evidence type | Direction | Claim tested | Key finding | Context | Confidence / limitations |
|---|---|---|---|---|---|---|---|
| 1 | This work (UniProt B8BAB0; InterPro/PANTHER/Pfam) | computational / database | qualifies | What subfamily is B8BAB0? | Signal peptide 1–21; BURP domain 384–595 (Pfam PF03181); InterPro IPR051897 "PG-associated BURP"; PANTHER PTHR31458:SF2 "Polygalacturonase 1 beta-like protein 2" | O. sativa indica, in silico | High; independent classifiers concur |
| 2 | This work (BLOSUM62 Needleman–Wunsch) | structural/evolutionary (computational) | refutes | Is B8BAB0 RAFTIN-like or PG1β-like? | 46.5% id to tomato PG1β (Q40161, score +1250); 44.9% to AroGP3 (P93218, +1163); only 28.2% to wheat RAFTIN1A (Q70KG5, NEGATIVE score −567) and 20–27% to other rice BURPs (negative scores) | full-length alignment | High; classic PG1β ortholog-level identity vs BURP-domain-only similarity to RAFTIN |
| 3 | 1392611 (Zheng 1992) | direct characterization | refutes/qualifies | What is the PG1β β-subunit repeat & function? | β-subunit is a fruit cell-wall glycoprotein "composed almost entirely of the novel 14-aa motif FTNYGxxGNGGxxx"; associates with & regulates catalytic PG2 | tomato fruit | High; B8BAB0 carries this diagnostic repeat (GFTNYGESGN, TFAGYGKNGN, ~37% of protein) |
| 4 | 14612572 (Wang 2003) | mutant phenotype / localization | competing | Do BURP proteins act in pollen? | RAFTIN (anther-specific, Ubisch-body/exine, essential for pollen) — but "occurrence of RAFTIN per se is limited to cereals," a distinct BURP clade | rice/wheat anther | High; establishes RAFTIN as the reproductive precedent — B8BAB0 is NOT in this clade (evidence #2) |
| 5 | 24237159 (Liu 2014) | mutant phenotype | refutes | Function of the rice PG1β-like clade | OsBURP16 (rice PG1β β-subunit): stress-inducible; overexpression ↓pectin, ↓cell adhesion, ↑abiotic-stress sensitivity | rice, whole plant | High; nearest functional analog of B8BAB0's clade — cell-wall/pectin, not pollen |
| 6 | 26106400 (Park 2015) | mutant phenotype / localization | qualifies | Function of PG1β-like proteins | AtPGL3 (Arabidopsis PG1β-like) localizes to cell wall, promotes cell enlargement; BURP domain required | Arabidopsis leaf | High; cell-wall function, not reproductive |
| 7 | 11554479 (Brummell 2001) | review | qualifies | Role of PG β-subunit | β-subunit protein "plays a role in limiting pectin solubilization"; affects middle lamella / cell adhesion | tomato fruit | Medium (review); mechanistic orientation |
| 8 | 12232274 (Zheng 1994) | expression / direct | qualifies | Where is PG1β expressed? | β-subunit fruit-predominant but also low levels in root, leaf, flower; developmentally (not ethylene) regulated | tomato | Medium; shows flower expression ≠ pollen-maturation function |
| 9 | 19363683 (Ding 2009) | expression / phylogenetic | refutes/qualifies | Which rice BURP clade is flower-predominant? | 17 OsBURP genes, 7 subfamilies; flower-predominant genes = monocot-specific "BURP VII" clade (RAFTIN-type); most others stress-induced | rice tissues | High; B8BAB0 (PG1β-like) is a different subfamily from the flower-predominant one |
| 10 | This work (AlphaFold AF-B8BAB0-F1 v6) | structural (computational) | qualifies | Is the BURP domain a real fold? | BURP domain (384–595) mean pLDDT 85.4 (well-folded); repeat region 77.6; N-term/signal 63.4; overall 74.6 | in silico model | Medium-High; confirms folded BURP domain, no catalytic/pollen inference |
GO Curation Implications (leads — require curator verification)
| GO aspect | Candidate term | Recommended action | Basis / evidence code |
|---|---|---|---|
| BP | pollen maturation (GO:0010152) — the predicted term | DO NOT ADD / REMOVE if present. Not supported by domain, subfamily, orthology, or structure. | Over-generalization from BURP superfamily; B8BAB0 is PG1β-like, not RAFTIN-like (#2,#4,#9) |
| CC | apoplast (GO:0048046) or extracellular region / cell wall (GO:0005576 / GO:0005618) | Candidate ISS lead (curator to verify) | Signal peptide + Glycoprotein keyword; PG1β/AtPGL3/GmRD22 all cell-wall/apoplast (#3,#6) |
| BP | plant-type cell wall organization (GO:0009664) or pectin catabolic process (GO:0045490) (as non-catalytic partner) | Weak ISS lead only; note β-subunit is non-catalytic | PG1β/OsBURP16 modulate pectin & cell adhesion (#3,#5,#7) |
| MF | polygalacturonase activity (GO:0004650) | DO NOT ADD. B8BAB0 is the non-catalytic β subunit; it is not a hydrolase. | #3,#5 |
| MF | protein binding | Avoid as a final term; the informative interaction (binding catalytic PG) is inferred from tomato only, not demonstrated for rice B8BAB0 | #3 |
Bottom line for the curator: the ProtNLM2 BP term should be treated as an incorrect computational prediction (over-annotation) and not propagated. If any homology-based annotation is desired, the defensible leads are CC = cell wall/apoplast and (weaker) BP = cell-wall/pectin organization, all as ISS/IEA-level leads requiring curator verification — never as experimental evidence.
Mechanistic Scope
- Immediate/direct gene-product function (what is testable): a secreted (signal peptide-bearing), glycosylated, low-complexity-repeat BURP protein of the PG1β β-subunit type. In the characterized paradigm the β-subunit is non-catalytic and acts as a cell-wall structural/regulatory partner that binds and modulates a catalytic polygalacturonase and limits pectin solubilization (tomato PG1; #3,#7).
- Downstream/indirect (NOT the direct molecular function): fruit softening, cell adhesion/expansion, and abiotic-stress tolerance are downstream cell-wall consequences seen for orthologs (#5,#6,#7). "Pollen maturation" is neither the direct activity nor an established downstream phenotype for this clade.
Conflicts and Alternatives
- Superfamily-vs-subfamily confusion (most likely source of the prediction): BURP is a large, functionally diverse superfamily. The reproductive-development association is real for the RAFTIN/BURP-VII clade (#4,#9) but does not transfer to the PG1β-like clade that B8BAB0 occupies (#2). ProtNLM2 likely over-generalized at the superfamily level.
- Architecture look-alike: B8BAB0's signal-peptide + long tandem-repeat + C-terminal BURP layout superficially
resembles RAFTIN, which could mislead an alignment-free predictor; but the repeat is the PG1β β-subunit
FTNYGxxGNGGmotif (#3), and quantitative alignment cleanly separates the two clades (#2). - Flower expression ≠ pollen maturation: tomato PG1β is detectable at low levels in flowers (#8); such expression, if it existed for B8BAB0, would not by itself justify the specific process term.
- Organism/paralog note: B8BAB0 is the indica ortholog of japonica OsBURP12 (Q6ZA27, 86.7% id); its paralog OsBURP14 (53.3% id) is more divergent. None of these is characterized as a pollen factor.
Knowledge Gaps
| Gap | What was checked | Why it matters | What would resolve it |
|---|---|---|---|
| B8BAB0-specific expression | UniProt, literature (19363683 family-level only); EBI Atlas & Bgee queried programmatically — no gene-level expression call returned (web-only atlases) | A curator could argue for tissue-specific pollen expression | RiceXPro / CREP / RAP-DB anther & pollen-stage RNA-seq for japonica ortholog RAP locus Os08g0380100 (OsBURP12) or indica BGIOSGA028580 |
| Direct in-planta function of OsBURP12 | No mutant/RNAi study found | The clade function is inferred from tomato/OsBURP16, not this gene | T-DNA/CRISPR knockout phenotyping in rice (cell-wall, fertility) |
| Interaction with a catalytic PG | Inferred from tomato PG1 only (#3) | Determines whether a "pectin/cell-wall" BP is defensible | Co-IP / pulldown of B8BAB0 with rice polygalacturonases |
Discriminating Tests
- Tissue expression: query anther-developmental-stage RNA-seq (RiceXPro/CREP) for BGIOSGA028580/OsBURP12. Pollen-maturation support would require strong, stage-specific anther/microspore expression and a loss-of-function pollen phenotype — the RAFTIN standard (#4).
- Reciprocal-best-hit / tree: build a BURP-domain ML tree placing B8BAB0 among RAFTIN (BURP VII) vs PG1β-like references; the alignment scores here (#2) predict it falls with PG1β/OsBURP16.
- Knockout phenotype: CRISPR knockout of OsBURP12; score seed set/pollen viability (RAFTIN-type) vs cell-wall/ pectin/stress traits (OsBURP16-type). The latter is expected.
Curation Leads (require curator verification)
- Action change: flag ProtNLM2
pollen maturation (GO:0010152)as NOT SUPPORTED (over-annotation) — do not add; remove/mark if present. Rationale: PG1β-like subfamily, not RAFTIN clade. - Candidate replacement leads (ISS/IEA, homology only):
- CC
apoplast (GO:0048046)/cell wall (GO:0005618)— from signal peptide + PG1β/AtPGL3 localization. - BP
plant-type cell wall organization (GO:0009664)— weak; note non-catalytic β-subunit role. - Candidate references + exact snippets to verify:
- PMID 1392611: "composed almost entirely of the novel 14-amino acid motif FTNYGxxGNGxxx" and "a cell wall protein that associates with, and apparently regulates, the catalytic PG2 polypeptides".
- PMID 24237159: "OsBURP16 belongs to the PG1β-like subfamily of BURP-family genes and encodes one putative PG1β subunit precursor in rice".
- PMID 26106400: "A C-terminal GFP fusion protein of AtPGL3 … localized to the cell wall".
- PMID 14612572: "occurrence RAFTIN per se is limited to cereals" (defines the separate reproductive clade).
- PMID 19363683: "all the genes of the BURP VII subfamily were predominantly expressed in flower organs" (the flower clade is distinct from B8BAB0's PG1β-like clade).
- Suggested curator questions: Is there anther/pollen-stage expression for BGIOSGA028580? Any rice OsBURP12 functional study? Is the review's prior action consistent with "no experimental support; homology → cell wall"?
- Suggested experiments: RNA-seq tissue profiling; CRISPR knockout fertility/cell-wall phenotyping; PG co-IP.
Limitations
- No wet-lab or B8BAB0-specific expression data were accessible; conclusions rest on sequence/structure/orthology computation plus primary literature on orthologs. This is sufficient to judge the sequence-based prediction but does not positively assign a validated function.
- Percent identities are from a custom global BLOSUM62 alignment (provenance retained), not a curated MSA; values are approximate but the clade separation (positive vs negative alignment scores) is robust.
Provenance: sequence analyses, alignments, motif detection, and AlphaFold pLDDT parsing were executed during this investigation (Iterations 1–2) and recorded in the knowledge state; an evidence-matrix CSV accompanies this report.