STRC OTOA Paralog Phase 1B Structure Alignment

Superimposing OTOA (AF DB Q7RTW8 v6) onto STRC (AF3 WT full-length) using the Cα atoms of all sequence-aligned residue pairs gives ARM-region RMSD 13.78 Å over 148 pairs, mini-STRC RMSD 9.85 Å over 362 pairs, global RMSD 35.8 Å over 1126 pairs. These are the sequence-forced superpositions; a structural aligner (TM-align/DALI, not run here) would find lower values by rearranging correspondences. Regardless of methodology, the headline is clear: STRC and OTOA share ancestry but have diverged in 3D fold. Two pharmacophore anchors (F1646, E1659) are identically conserved in OTOA; two (K1141, G1645) fall in gaps. Verdict: paralog at sequence level, NOT at fold level — the chimera avenue collapses; OTOA upregulation still plausible only if OTOA’s own fold happens to provide bundle-mechanical function despite structural divergence.

Method

  1. Load STRC structure from our AF3 WT CIF (job4-wildtype.cif, full-length 1–1775, covers ARM and pocket regions).
  2. Load OTOA AlphaFold DB structure v6 (AF-Q7RTW8-F1-model_v6.pdb, 1153 residues, global metric 84.1, 87.6% residues at confident+very high pLDDT).
  3. Validate structure-sequence correspondence: sample Cα residues at positions 1000/1600/1700 (STRC) and 500/1000 (OTOA) match UniProt sequence letters.
  4. Run local BLOSUM62 alignments on (a) ARM 1603–1770, (b) mini-STRC 700–1775, (c) global full length, against OTOA.
  5. Collect every sequence-aligned pair as (STRC resnum, OTOA resnum).
  6. Feed matched-pair Cα atoms into Biopython Superimposer — produces rigid-body rotation+translation minimising Cα RMSD.

No structural aligner (TM-align/DALI) used; this is the sequence-forced superposition, which is an upper bound on true structural RMSD.

Results

RegionSTRC spanPaired CαCα RMSDSeq identity at aligned
ARM repeats1603–177014813.78 Å33.8%
Mini-STRC700–17753629.85 Å29.8%
Global1–1775112635.82 Å28.3%

For reference, structurally homologous paralog pairs typically yield ≤4 Å Cα RMSD over conserved cores; distant homologs with shared fold 4–6 Å; unrelated or divergent folds >10 Å. The ARM-region value of 13.78 Å sits unambiguously in the “divergent fold” regime despite strong sequence conservation. The ARM aligned block (STRC 1603–1770 ↔ OTOA 977–1136) carries 34% identity yet occupies geometrically different 3D positions in the two predicted structures — classic sign of a paralog pair that evolved into different folds after gene duplication (precedent: many α/β-propeller vs α-solenoid paralog pairs).

Structure-sequence validation

PositionSTRC struct CαSTRC seqOTOA struct CαOTOA seq
STRC 1000LL
STRC 1600GG
STRC 1700FF
OTOA 500SS
OTOA 1000TT

All match. Structure files are correctly registered to UniProt numbering.

Pharmacophore anchors in OTOA

STRC residueSTRCOTOA equivalent (global aln)Conservation
K1141KgapSTRC-specific
G1645GgapSTRC-specific
F1646FF1014identical
E1659EE1027identical

F1646 and E1659 are identically conserved despite the ARM fold diverging — they are functionally critical residues whose side-chain chemistry has been preserved across paralogs. This is striking: in a region where the 3D fold has diverged enough to give 14 Å RMSD, two of our four anchor residues are still the same letter. The sequence-level conservation of E1659 adds an independent argument (beyond cross-species ortholog conservation already recorded in STRC Electrostatic Analysis E1659A) that this residue is load-bearing in both STRC and OTOA function.

Interpretation

Avenue (from parent note)Phase 1B verdict
1. Transcriptional upregulation of endogenous OTOA in OHCsWeakened but alive. If OTOA is merely ectopically expressed in OHCs, its natural fold must substitute for STRC’s bundle-mechanics role. With 14 Å ARM RMSD the two domains do not share 3D geometry → OTOA cannot directly replace STRC’s TMEM145-docking ARM architecture. Residual rescue possible only through OTOA’s native function (TM-side anchoring).
2. Engineered OTOA-HTC chimeraKilled as a simple graft. Structural divergence means a STRC-HTC surface grafted onto OTOA’s C-terminal will not sit at the same 3D location relative to the rest of the protein. Chimera design would require de novo interface engineering + AF3 iteration — an order of magnitude more work than originally scoped. Do not pursue without prior proof-of-concept in a simpler paralog pair.
3. Small-molecule OTOA activatorUnchanged. Sequence/structure tell nothing about OTOA’s promoter druggability.
NEW observationThe identical conservation of F1646/E1659 across STRC↔OTOA is itself a usable biomarker: any cross-species variant at these positions is predicted pathogenic for both proteins.

Caveats and limitations

  • Sequence-forced superposition is an upper bound. A proper structural aligner (TM-align/DALI) would find different correspondences. If the real TM-align RMSD drops to ≤6 Å over an ARM sub-core, the chimera verdict could be revised. A TM-align rerun is a cheap follow-up.
  • AF predictions, not experimental structures. AF3 for STRC is high pTM (0.86); AF DB for OTOA is high global metric (84). Both are credible, but neither is a crystal structure. An incorrect AF prediction of either protein would invalidate the RMSD.
  • Anchor conservation does not imply functional substitution. F1646/E1659 being identical in OTOA says these residues have important chemistry in both proteins — but not that they are positioned for the same role.
  • OTOA’s natural function in the cochlea still matters. Even without STRC-level fold homology, ectopic OTOA expression in OHCs may provide partial TM anchoring (its native function in supporting cells). That is a biological question, not a structural one.

Next steps

  1. TM-align or DALI reconstruction. Cost: minimal. Re-evaluates whether ARM and OTOA C-terminal share any sub-core.
  2. Secondary-structure composition check. Compute α-helix / β-sheet fraction for STRC 1603–1770 vs OTOA 977–1136. Different SSE mix → confirms fold divergence without ambiguity.
  3. Revised parent note. Downgrade Avenue 2 (chimera) from “feasible with engineering” to “speculative; requires fresh structure-first design.” Keep Avenue 1 on the active list but flag the fold-divergence caveat.

Replication

cd ~/STRC/models
# STRC structure: ~/Sites/site-strc-egor-lol/public/models/job4-wildtype.cif
# OTOA structure: downloaded at runtime
curl -sSL -o /tmp/af_otoa.pdb \
  https://alphafold.ebi.ac.uk/files/AF-Q7RTW8-F1-model_v6.pdb
/opt/miniconda3/bin/python3 otoa_phase1b_structure_alignment.py

Files / Models

  • ~/STRC/models/otoa_phase1b_structure_alignment.py — structure load, sequence alignment, Biopython Superimposer
  • ~/STRC/models/otoa_phase1b_structure_alignment.json — per-region RMSD + anchor residues + sample pairs
  • /tmp/af_otoa.pdb — AlphaFold DB Q7RTW8 v6 (regenerable via curl above)

Connections