glo (Q8INJ6): evidence and ProtNLM claim review

Glorund is an RNA-binding regulator of translation and RNA processing in Drosophila oogenesis. The Q8INJ6 isoform-C record is a short 179-residue RRM-containing product, whereas gene-level experiments establish repression of nanos translation, association with Hrp48 and Half-pint, and roles in oocyte RNA localization. Which of these activities and localizations are retained by this short isoform remains unresolved.

Exact input: Q8INJ6, 179 residues. The accession was fetched explicitly with the gene-directory alias; no canonical-sequence substitution is made.

Raw emitted predictions: glo-predictions-source.json. Source features: glo-uniprot.txt, with an exact extraction in glo-sequence-evidence.json.

Sequence and domain evidence

These are sequence/domain observations or explicitly named feature predictions, not measurements of biological function. ARBA assertions and ProtNLM-derived UniProt names are not counted as validation.

DR   InterPro; IPR050666; ESRP.
DR   InterPro; IPR012677; Nucleotide-bd_a/b_plait_sf.
DR   InterPro; IPR035979; RBD_domain_sf.
DR   InterPro; IPR000504; RRM_dom.
FT   DOMAIN          49..129
FT                   /note="RRM"
FT                   /evidence="ECO:0000259|PROSITE:PS50102"

ProtNLM claims

The snapshot emits names and location/keyword statements, with no GO or EC prediction for this target. Each actual statement is assessed below; no GO term has been substituted for it. Categories follow the function-prediction rubric, with nonspecific “uncharacterized” names marked UNC because they contain no testable function. CNN records an independently supported existing annotation; it does not assert a particular training-set composition.

Kind	Verbatim emitted statement	Assessment	Evidence and limitation
Name	RRM domain-containing protein	CNN	IPR000504 and the domain feature at residues 49-129 independently establish the RRM in the exact 179-residue sequence. This limited domain claim remains sound despite uncertainty about full-length Glorund functions.
Location	Nucleoplasm (SL-0190)	UNC	PMID:19013444 discusses nuclear and cytoplasmic Glorund, but the localization is not mapped to the short 179-residue Q8INJ6 isoform. A nuclear-splicing role at the locus does not independently establish nucleoplasmic residence of this product.
Location	Nucleus (SL-0191)	UNC	Nuclear Glorund is reported in the literature, but the exact Q8INJ6 isoform is short and requires construct mapping. Broad nuclear residence of this product remains unresolved rather than refuted.

Literature evidence

PMID:19013444(https://pubmed.ncbi.nlm.nih.gov/19013444/): “whereas Glo and Hrp48 are both nuclear and cytoplasmic”
PMID:19013444(https://pubmed.ncbi.nlm.nih.gov/19013444/): “Here we show that Glorund is part of a complex containing the hnRNP protein Hrp48 and the splicing factor Half-pint”
PMID:16516833(https://pubmed.ncbi.nlm.nih.gov/16516833/): “Here we identify a Drosophila hnRNP, Glorund, that interacts specifically with stem-loop III.”

Annotation decisions

GO:0003676 nucleic acid binding (IEA): MODIFY. The exact product retains an RRM at residues 49-129 and belongs to the experimentally characterized RNA-binding Glorund locus. RNA binding is a reasonable domain-level inference without transferring a specific mRNA target.
GO:0003723 RNA binding (IEA): ACCEPT. The 179-residue product contains an intact annotated RRM at residues 49-129. This supports broad RNA binding but does not identify its target RNAs or establish full-length Glorund activity.
GO:0003729 mRNA binding (ISS): UNDECIDED. PMID:10908586 surveys RNA-binding proteins at gene/family level; the exact Q8INJ6 product has only a short RRM-containing architecture and no isoform-mapped mRNA binding assay is established.
GO:0005515 protein binding (IPI): UNDECIDED. PMID:19013444 identifies a Glorund-Hrp48-Half-pint complex. The 179-residue isoform needs construct mapping before these gene-level interactions can be assigned to it.
GO:0005515 protein binding (IPI): UNDECIDED. The full-text complex experiments in PMID:19013444 establish Glorund-associated interactions, but do not establish retention of the interaction surfaces in the exact 179-residue product.
GO:0006417 regulation of translation (IMP): UNDECIDED. PMID:16516833 demonstrates Glorund-dependent nanos translational repression. Q8INJ6 contains a short RRM-bearing sequence and requires construct or isoform mapping before that activity is transferred.
GO:0007311 maternal specification of dorsal/ventral axis, oocyte, germ-line encoded (IMP): UNDECIDED. PMID:19013444 establishes Glorund roles in oocyte patterning; the responsible experimental products are not mapped to Q8INJ6 in the inspected evidence.
GO:0032991 protein-containing complex (IPI): UNDECIDED. The Glorund-Hrp48-Half-pint complex in PMID:19013444 is real gene-level evidence; incorporation of the 179-residue isoform is unestablished.
GO:0032991 protein-containing complex (IPI): UNDECIDED. The Glorund-Hrp48-Half-pint complex in PMID:19013444 is real gene-level evidence; incorporation of the 179-residue isoform is unestablished.
GO:0042060 wound healing (HMP): UNDECIDED. PMID:19884309 is an epithelial-repair genetic screen and the cache is abstract-only. Neither the precise experiment nor isoform-specific contribution can be verified from that source.
GO:0045451 germ plasm oskar mRNA localization (IMP): UNDECIDED. PMID:19013444 establishes a Glorund-dependent mRNA-localization phenotype; an intact RRM alone does not show that the 179-residue isoform executes this process.
GO:0051276 chromosome organization (IMP): UNDECIDED. PMID:19013444 connects Glorund complexes and ovarian-tumor splicing to chromosome organization. The phenotype should not be transferred to the short isoform without construct mapping.
GO:0060810 intracellular mRNA localization involved in pattern specification process (IMP): UNDECIDED. The locus-level data in PMID:19013444 support Glorund biology, while the sequence represented by Q8INJ6 requires specific mapping to the assayed protein.

Research provenance

Genuine external literature research is requested through the repository Falcon wrapper, with perplexity-lite configured as fallback. Provider output is retained separately as glo-deep-research-<provider>.md; its source leads are checked against the underlying publications and exact sequence record.

2026-10-21 — Full re-review, independent of the ProtNLM2 predictions

This re-review did not use ProtNLM2 output. Sources were the UniProt record, the GOA rows, PANTHER/PAINT data, the falcon deep-research file, and seven newly cached full-text Glorund papers.

Identity of the benchmark accession

Q8INJ6 is Glorund isoform C, 179 aa, TrEMBL, from FlyBase transcript CG6946-RC. It is not the canonical product. The long product is Q9VGH5, 586 aa, TrEMBL, covering isoforms A/B/D. Q8IG99 (SD14463p, 359 aa) is a cDNA covering the C-terminal half. None of the glo entries is Swiss-Prot reviewed.
glo-bioinformatics/compare_isoforms_rereview.py produced RESULTS-rereview.md and results-rereview.json. Residues 1-168 of Q8INJ6 are identical to Q9VGH5. Q8INJ6 then ends in the unique tail ILRYSAARLLK.
Using the qRRM boundaries of PMID:35699205 (qRRM1 45-141, qRRM2 142-234, qRRM3 475-562), isoform C keeps all of qRRM1, only residues 142-168 of qRRM2, and nothing of the Gly/Asn-rich region or qRRM3.
Functional consequence: Warden et al. 2023 PMID:37427795 and ["at least two qRRMs are required for TCE recognition and translational repression of nos in vivo"]. Isoform C alone is therefore not expected to repress nanos. Its qRRM1 should still bind G-tract RNA PMID:28380354. It also lacks most of qRRM2, the dFMRP-binding domain PMID:35699205.
GOA attaches the FlyBase gene-level annotations to both Q8INJ6 and Q9VGH5. Only Q9VGH5 carries the PAINT IBA rows: RNA binding, nucleoplasm, ribonucleoprotein complex and GO:0043484, from PTN000360053 and PTN000359854. I reviewed the FlyBase rows as claims about the glo gene and stated in each reason whether the claim can apply to the isoform-C product.

Evolutionary placement

PANTHER 19.0 places glo (Q9VGH5) in PTHR13976:SF76 "AT27789P", with C. elegans hrpf-2 (Q8MXR2) and plant proteins (At3g20890, two poplar genes). It is not in the vertebrate hnRNP F/H subfamilies (SF32 HNRNPF, SF33 HNRNPH2, SF36 HNRNPH3, SF60 HNRNPH1) or the ESRP subfamilies (SF30, SF37). Fly fusilli, the ESRP ortholog, is in SF25.
The name "hnRNP F/H homolog" therefore reflects shared qRRM architecture and conserved G-tract-binding residues PMID:28380354. It does not mean subfamily-level orthology. I made no subfamily-specific transfer from human hnRNP F/H, such as splicing of particular targets. The only phylogenetic support used is the PAINT node PTN000360053, which spans the clade and gives RNA binding.
The family review (interpro/panther/PTHR13976/PTHR13976-review.yaml) lists Q8INJ6 as a representative of one lumped "ESRP, hnRNP H/F and related" subfamily. That conflicts with PANTHER, which separates glo (SF76) from ESRP and hnRNP F/H. The family review also uses the isoform accession rather than Q9VGH5. Both points are recorded here; the family file was not edited.

Annotation decisions (gene-level claims)

GO:0003676 nucleic acid binding (IEA): MODIFY to RNA binding.
GO:0003723 RNA binding (IEA): ACCEPT. qRRM1, which isoform C retains, binds RNA directly.
GO:0003729 mRNA binding (ISS, PMID:10908586): ACCEPT on direct evidence (PMID:16516833, PMID:28380354, PMID:37427795). The cited survey does not mention Glo (reference marked MISCITED).
GO:0005515 protein binding x2 (Hrp48, Hfp): REMOVE under the protein-binding policy. The interactions are real and RNase-resistant.
GO:0006417 regulation of translation: MODIFY to GO:0017148 negative regulation of translation.
GO:0007311 maternal D/V specification, GO:0060810 mRNA localization in pattern specification, and GO:0045451 oskar mRNA localization: KEEP_AS_NON_CORE. These are genuine germline requirements but downstream of otu splicing and G-tract-mode RNA regulation.
GO:0051276 chromosome organization: MARK_AS_OVER_ANNOTATED. Glo does none of the chromosome-organization work; Otu-104 partly rescues the phenotype.
GO:0032991 protein-containing complex x2: KEEP_AS_NON_CORE.
GO:0042060 wound healing (HMP, screen): UNDECIDED. Only the abstract is available, and it does not mention glo.
NEW: GO:0003730 mRNA 3'-UTR binding, GO:0000900 mRNA regulatory element binding translation repressor activity, GO:0005634 nucleus, and GO:0005737 cytoplasm. All are gene-level and stated not to be demonstrated for isoform C.
I did not propose terms for the lipid-storage RNAi (PMID:33537463, PMID:41837177) or testis overexpression (PMID:39092172) phenotypes. They are organismal or downstream phenotypes.

Relative to review next

Q9VGH5, the long canonical product of the same gene (same_gene_canonical), has no review in the repo. Its folder would be genes/DROME/glo__Q9VGH5. That review should cover the IBA rows (RNA binding, nucleoplasm GO:0005654, ribonucleoprotein complex GO:1990904, GO:0043484 regulation of RNA splicing) that are absent from Q8INJ6.