2025 Zhao — Deciphering enhancers of hearing loss genes for efficient and targeted gene therapy of hereditary deafness
ARBITER workflow engineers B8 OHC-specific enhancer (706 bp, E1P3×2+E2P2×2+E2P3×2) that restores hearing in Slc26a5-KO mice with ~100% OHC transduction and zero ectopic expression.
Citation
Zhao S, Yang Q, Yu Z, Chu C, Dai S, Li H, Diao M, Feng L, Ke J, Xue Y, Zhou Q, Liu Y, Ma H, Lin C-P, Yao Y-G, Zhong G. “Deciphering enhancers of hearing loss genes for efficient and targeted gene therapy of hereditary deafness.” Neuron. 2025 May 21;113(10):1579-1596.e5. DOI: 10.1016/j.neuron.2025.03.023. PMID: 40262614.
NOTE: Paper is in Neuron, NOT Cell. Year 2025, NOT 2024. Script phantom cite “Yoshimura 2018” is entirely wrong — this Zhao 2025 Neuron paper is the correct and only source for B8/ARBITER.
TL;DR
ARBITER (AAV-reporter-based in vivo transcriptional enhancer reconstruction) dissects conserved non-coding elements (CNEs) near hearing loss gene loci to engineer OHC-specific synthetic enhancers. Applied to Slc26a5 (prestin), ARBITER identified E1 (247 bp) and E2 (525 bp) CNEs in intronic regions, dissected them into key conserved modules, and reassembled them into B8 — a synthetic enhancer achieving ~100% OHC transduction with zero IHC, vestibular, brain, heart, or liver off-target expression. B8-Slc26a5 gene therapy fully restored ABR thresholds in Slc26a5-KO mice at high dose (5e10 gc/mouse).
Numbers that matter
| Parameter | Value | Units | Source (page/fig/table) | Uncertainty |
|---|---|---|---|---|
| Slc26a5-E1 size | 247 | bp | Fig 2A, text p.1582 | exact genomic sequence |
| Slc26a5-E2 size | 525 | bp | Fig 2A, text p.1582 | exact genomic sequence |
| E1P3 module (most conserved E1 fragment) | 93 | bp | text p.1586, Fig S6A | exact |
| E2P2 module (most conserved E2 fragment) | 132 | bp | text p.1586, Fig S6B | exact |
| E2P3 module (second conserved E2 fragment) | 128 | bp | text p.1586, Fig S6B | exact |
| B8 composition | E1P3×2 + E2P2×2 + E2P3×2 | — | text p.1587, Fig 6A | exact module identity |
| B8 calculated size | 706 | bp | back-calculation: 2(93)+2(132)+2(128) | no explicit bp stated in text — back-calc from module sizes |
| B8 OHC transduction efficiency | ~100% | % | Fig 6C (apical/middle/basal) | near-complete; exact % read from bar graph |
| B8 ectopic expression | 0 | — | Fig 6D, 6E, 6F, 6G | zero in IHC, vestibular, brain, heart, liver |
| Low dose gene therapy (B8-Slc26a5) | 1e10 | gc/mouse | Fig 7, text p.1588 | partial ABR rescue (4/8 mice) |
| High dose gene therapy (B8-Slc26a5) | 5e10 | gc/mouse | Fig 7, text p.1588 | full ABR rescue to WT level (8/11 mice) |
| AAV capsid used | AAV-ie (neonatal); AAV2 (adult) | — | text p.1583, p.1588 | |
| Slc26a5-E3 size (non-conserved, negative control) | 308 | bp | Fig 2A | negative result |
| Slc26a5-E4 size (non-conserved, negative control) | 489 | bp | Fig 2A | negative result |
| All synthetic reporter sequences | See Table S2 | — | Table S2 (supplementary Excel file) | full sequence list for At-, Sl-, My-, A-, B-series |
| CAG promoter OHC transduction | ~100% | % | Fig S1B | also infects IHC + brain — no specificity |
| CAG promoter brain infection | yes (expansive) | — | Fig S1E | disqualifying for clinical use |
B8 SIZE (resolved 2026-04-24): Main text does not state an explicit total bp, but the composition (Fig 6A, text p.1587) and module sizes (Fig S6A/B, text p.1586) are unambiguous: E1P3×2 + E2P2×2 + E2P3×2 with E1P3=93, E2P2=132, E2P3=128 → contiguous back-calc 706 bp. Scripts adopted 706 bp 2026-04-24; prior 587 bp (no literature backing) retracted per AGENTS.md §0c.4. Table S2 (SI Excel, “AAV reporter vectors carrying different synthetic enhancers — This paper, see Table S2”) gives the exact cloned sequence including KpnI/XbaI restriction-site linkers (Key Resources Table); treated as a cloning-QC gate before GMP submission, not as a design blocker. Re-run of ultramini_vector_cpg_audit.py with 706 bp: full vector 3565/4700 bp (75.9% of AAV capacity), 48 CpG; ultra_mini_promoter_shortlist.py: B8+WPRE3 still top candidate with 891 bp spare.
Method essentials
- Mouse model: neonatal C57BL/6 + Slc26a5-KO
- Injection: round window injection (RWI) at P2-P3 neonatal; posterior semicircular canal (PSCC) in adults and tree shrews
- Reporter: AAV-ie-[enhancer]-MinP-nls-mNeonGreen
- Readout: immunofluorescence (transduction efficiency), ABR + DPOAE (hearing function), SEM (stereocilia morphology)
- CNE identification: ATAC-seq (GSE181311, GSE150000), PhyloP conservation score >1.0, UCSC Genome Browser
- ARBITER validated first on Atoh1 enhancers, then applied to Slc26a5 and Myo7a
Limitations
- B8 characterized only in mouse cochlea and tree shrew — no human OHC validation
- B8 size in bp is not explicitly stated in main text; Table S2 (supplementary) required for exact cloned sequence
- No long-term expression stability data (epigenetic silencing risk unknown)
- Slc26a5-KO model: prestin loss causes OHC motility defect but cells survive; STRC model (null allele) involves structural OHC damage that may differ
- ABR rescue frequency-dependent: strongest recovery at 16 kHz; broader frequency rescue data limited
Relevance to STRC
B8 is the selected OHC-specific enhancer for the h03 Mini-STRC single-vector AAV construct. The ARBITER paper is the definitive source for:
- B8 composition and module identity
- B8 OHC-exclusivity (zero ectopic)
- B8 safety (no ABR threshold shift in WT mice)
The 587 bp value in h03 scripts (ultra_mini_promoter_shortlist.py and ultramini_vector_cpg_audit.py) does not match the 706 bp back-calculation from this paper. The script docstring phantom cite “Yoshimura 2018 ARBITER” must be corrected to cite this paper.
Connections
[part-of]aav-cochlear-delivery[supports]STRC B8 Enhancer Selection[applies]h03 hub[see-also]STRC Ultra-Mini Promoter Shortlist