AIGR Deep Research — S. pombe iwr1 (O13951) core-function hypothesis

Hypothesis under review: Schizosaccharomyces pombe Iwr1 is a conserved RNA polymerase II
nuclear-import adaptor that binds assembled Pol II and presents its own NLS to karyopherin-α.

Focus type: core_function · Gene: iwr1 / SPAC23H4.08 · UniProt: O13951 (277 aa)

Executive Judgment

Verdict: Partially supported (supported by orthology; not experimentally demonstrated in S. pombe).

The described activity — an Iwr1 protein that binds assembled Pol II and presents its own
N-terminal bipartite NLS to karyopherin-α to drive Pol II nuclear import — is a bona fide,
directly demonstrated core molecular function of the Saccharomyces cerevisiae ortholog
(Czeko et al., 2011, PMID 21504834), and the authors explicitly report it is "apparently
conserved from yeast to human." For the S. pombe target specifically, the assignment is
entirely inference-based: every GO annotation on O13951 is electronic/phylogenetic
(IEA/ISO/IBA), the molecular-function slot is ND (no data), and no primary S. pombe
experimental study of iwr1 exists in the literature. Sequence analysis I ran supports
transferability — O13951 carries the Iwr1 Pfam domain (PF08574) and conserves the N-terminal
basic (NLS-like) clusters and acidic C-terminal tail seen in budding-yeast Iwr1 — but the NLS /
karyopherin-α interaction and Pol II binding have not been tested in fission yeast.

Most important caveats. (1) The "presents its own NLS to karyopherin-α" clause is a
S. cerevisiae mechanistic detail, transferred by conservation, not verified in S. pombe.
(2) Iwr1 is non-essential and Pol II import can proceed by Iwr1-independent routes
(PMID 26455955), so it is a facilitating adaptor, not the obligate importer. (3) Budding-yeast
Iwr1 has additional documented roles (elongation, Pol II disassembly from chromatin, genome
stability, gene-specific transcription); the import-adaptor role is the best-supported direct
function, whereas transcription phenotypes are largely downstream. (4) Ortholog identity is
low (~23% to S. cerevisiae, ~20% to human SLC7A6OS); orthology rests on the shared Iwr1
Pfam domain (PF08574) and positionally conserved N-terminal basic clusters rather than high
sequence identity.

Supporting feasibility checks I ran this iteration. The AlphaFold model of O13951 is largely
low-confidence/disordered (mean pLDDT 64.6) but the N-terminal NLS cluster (res 5–15) is
well-ordered (pLDDT ≈91) — structurally consistent with an intrinsically disordered Pol II-cleft
adaptor carrying a defined, accessible NLS. S. pombe also encodes the required importin-α
receptors (imp1/O94374, cut15/O14063), so the proposed karyopherin-α–dependent mechanism is
feasible in this organism (though the Iwr1–importin-α interaction itself is untested here).

Evidence Matrix

GO Curation Implications (leads — require curator verification)

• BP GO:0006606 "protein import into nucleus" — RETAIN. Well-supported by orthology
  (ISO from SGD IWR1, IBA, IEA) and consistent with conserved architecture. It is appropriately
  general; there is no granular "RNA polymerase II complex nuclear import" GO term, so 0006606 is
  the correct level. Evidence code should remain non-experimental (ISO/IBA) for S. pombe.
• MF is currently ND — a candidate ADD (lead). The hypothesis implies a definable molecular
  function as an import adaptor. Best-supported candidate MF terms, by orthology only:
• GO:0030674 "protein-macromolecule adaptor activity" — captures bridging Pol II cargo to
  karyopherin-α; the most informative MF consistent with the mechanism.
• GO:0000993 "RNA polymerase II complex binding" — captures the direct Pol II association.
• Avoid "protein binding" (GO:0005515) as a terminal recommendation; the above are more informative.
• Because S. pombe evidence is inference-only, any MF addition should carry ISO/IBA evidence,
  not experimental codes.
• CC GO:0005634 (nucleus) + GO:0005737 (cytoplasm) — RETAIN. Consistent with a shuttling
  adaptor and with the S. pombe ORFeome localization resource (PMID 16823372, verify the
  iwr1-specific call).
• Do not over-state. Avoid annotating iwr1 as the obligate Pol II importer or asserting a
  verified karyopherin-α interaction in S. pombe; the import pathway is redundant (PMID 26455955).

GO Decision Table (leads — require curator verification)

All S. pombe MF/BP/CC calls should carry non-experimental (ISO/IBA) evidence — no
experimental (EXP/IDA/IPI) code is justified, as no direct fission-yeast assay exists.

Mechanistic Scope

Direct molecular function under test: cytoplasmic binding of the assembled 12-subunit Pol II
in its cleft, and presentation of Iwr1's own N-terminal bipartite NLS to karyopherin-α
(importin-α; S. pombe Cut15/Imp1 family) to mediate Pol II nuclear import, followed by Iwr1
displacement, export and recycling. This is a transcription-independent transport/adaptor
activity.

Downstream / non-core effects (should not be conflated with the core function):
gene-specific transcription changes, NC2/TFIIA genetic bypass, elongation and Pol II
chromatin-disassembly roles, genome instability and genotoxin sensitivity in null mutants.
These are documented for budding yeast and are largely consequences of (or roles adjacent to)
the import/assembly function rather than the primary molecular activity.

Conflicts and Alternatives

• Organism gap (principal issue): all direct evidence is S. cerevisiae/human-family; the
  S. pombe claim is orthology transfer. This is the central caveat, not a contradiction.
• Redundancy: Iwr1-independent import pathways exist (PMID 26455955); "adaptor" is accurate,
  "the importer" is too strong.
• Multifunctionality: Gómez-Navarro 2017 (PMID 28258010) supports additional direct roles
  (elongation, Pol II turnover), so "nuclear-import adaptor" may be one of several direct
  functions rather than the sole one — though it remains the best-characterized primary activity.
• Pol specificity: Minaker 2013 (PMID 23267056) reports iwr1 mutant Pol III localization
  defects too, mildly complicating the strict "Pol II-specific" clause.
• Paralog/family note: family is IWR1/SLC7A6OS (human SLC7A6OS); PomBase lists iwr1 as
  "predominantly single copy (one to one)" with no S. pombe paralog, so paralog
  over-annotation is unlikely here.
• Phenotype direction conflict: S. pombe iwr1Δ shows resistance to MMS/HU in genome-wide
  screens, whereas S. cerevisiae iwr1Δ is sensitive to genotoxins (PMID 28258010). Likely an
  organism difference or high-throughput screen artifact; either way these DNA-damage phenotypes
  are peripheral to the core import-adaptor function and should not drive the core-function call.
• Taxonomic-scope nuance: PomBase tags iwr1 both "conserved in eukaryotes" and "conserved in
  fungi only"; a clear 1:1 ortholog is fungal, while the broader Pfam family reaches metazoa
  (human SLC7A6OS) only at ~20% identity. The seed's "conserved… yeast to human" is defensible at
  the family level but weak at the sequence-identity level.

Knowledge Gaps

1. No direct S. pombe assay of Iwr1–Pol II binding or Iwr1–karyopherin-α (Cut15/Imp1)
   interaction. Checked: PubMed (no iwr1 fission-yeast primary papers), UniProt (by-similarity),
   QuickGO (no EXP codes). Matters: the entire core-function call is inference. Resolve:
   co-IP/affinity capture of Pol II with tagged Iwr1 in S. pombe; NLS-deletion import assay.
2. Functional NLS not verified in O13951. Checked: sequence — basic clusters present in the
   conserved position. Matters: the "own NLS" clause is the crux of the hypothesis. Resolve:
   mutate res 7–11/34–38 and assay Rpb1 nuclear localization; test suppression by Rpb3-NLS fusion.
3. iwr1-specific localization in S. pombe — partially resolved. Checked: PomBase
   records CC cytoplasm + nucleus and the descriptor "nucleocytoplasmic shuttling protein," and
   flags characterisation_status = "biological role inferred." Remaining: these lack an
   experimental evidence code tied to a dedicated iwr1 study; confirm the ORFeome (PMID 16823372)
   call and whether shuttling was directly observed for iwr1 in S. pombe.
4. Essentiality / redundancy in S. pombe unquantified. Resolve: iwr1Δ phenotype and
   Rpb1-GFP localization in fission yeast.

Discriminating Tests

1. Iwr1–Pol II co-purification in S. pombe (TAP/FLAG-Iwr1 → MS): confirms cargo binding.
2. NLS-swap / NLS-deletion import assay: delete O13951 N-terminal basic clusters, score
   Rpb1-GFP nuclear import; rescue by fusing a heterologous NLS to Rpb3 (mirrors PMID 23267056).
3. Iwr1–karyopherin-α (Cut15) two-hybrid / pulldown in S. pombe.
4. Structure prediction of the O13951–Pol II cleft interface (AlphaFold-Multimer of Iwr1 with
   Rpb1/Rpb2) to test cleft-binding conservation in silico.
5. Cross-complementation: does S. pombe iwr1 rescue S. cerevisiae iwr1Δ import defect?

Curation Leads (require curator verification)

• Action: Keep BP GO:0006606; consider adding an MF (GO:0030674 protein-macromolecule
  adaptor activity and/or GO:0000993 RNA polymerase II complex binding) with ISO/IBA evidence,
  filling the current ND slot. Keep CC nucleus + cytoplasm.
• Reference to cite for the mechanism (ortholog): PMID 21504834 — snippet to verify:
  "Iwr1 then uses an N-terminal bipartite nuclear localization signal that is recognized by
  karyopherin α to direct Pol II nuclear import" and "Iwr1 function is Pol II specific,
  transcription independent, and apparently conserved from yeast to human."
• Qualifier reference (redundancy): PMID 26455955 — establish that import is not
  Iwr1-obligate; annotate as adaptor/facilitator, not sole importer.
• Localization support (fission yeast): PMID 16823372 (verify iwr1 call).
• Suggested question for curator: Should the S. pombe MF/BP additions be evidence-coded
  ISO/IBA (recommended) given the absence of any S. pombe experimental data?
• Suggested experiments: the discriminating tests above, especially the NLS-deletion import
  assay and Iwr1–Pol II co-IP in S. pombe.

Provenance: UniProt REST (O13951, Q07532), QuickGO annotation API, and an executed
sequence-conservation comparison (Iwr1 domain PF08574; N-terminal basic clusters; acidic C-tail)
were run during this review. All S. pombe-level functional statements are inference from
orthology; no S. pombe experimental primary literature was found.

Artifacts

• OpenScientist final report
• OpenScientist final report