STRC RBM24 Regulatory Hypothesis

Core claim

RBM24 is an RNA-binding protein that directly regulates STRC mRNA via alternative splicing. If STRC expression in a patient is reduced (not absent) due to a regulatory or hypomorphic mutation, upregulating RBM24 activity or correcting its splicing targets could restore endogenous STRC to therapeutic levels — without delivering any exogenous gene. No AAV, no size constraint, no immune response.

Source

2026-04-17-sun-rbm24-strc-splicing — Sun et al. 2026 (PNAS, PMID 41973913): deletion of Rbm24 exon 4 causes stereocilia disorganization and hair cell loss in mice. STRC identified as a direct RBM24 splicing target.

Why this matters

Every other hypothesis starts from: “STRC is broken, deliver a replacement.” This one asks: “Is STRC actually silent, or just underexpressed? Can we turn up what’s already there?”

For patients with hypomorphic STRC mutations (partial-function alleles, regulatory variants, or compound heterozygotes where one allele retains partial function — like Misha’s VUS c.4976A>C), there may be residual STRC protein that could be boosted. Even a 2–3× increase in expression could cross the threshold for functional hair bundle mechanics.

Mechanism candidates

1. RBM24 overexpression via AAV — deliver RBM24 (not STRC) to OHCs. Smaller payload (~1.5 kb CDS), fits easily in single AAV. RBM24 then upregulates endogenous STRC splicing.
2. Small molecule RBM24 activator — if RBM24 activity is druggable, a systemic or intracochlear small molecule could boost STRC expression without any gene therapy.
3. ASO targeting RBM24 splicing — antisense oligonucleotide to promote the exon 4-including isoform of RBM24 (the functional one), delivered intracochlearly.

Patient stratification — when this applies

This hypothesis only works for a subset of DFNB16 patients:

• Hypomorphic STRC mutations (partial protein, not null)
• Regulatory variants affecting STRC transcription/splicing
• Compound heterozygotes with one functional allele (could be boosted)

It does NOT apply to patients with two null alleles (complete STRC loss) — there is nothing to boost.

Misha: compound heterozygous (paternal deletion + maternal c.4976A>C VUS). The maternal allele produces partial STRC (E1659A, reduced TM binding but not null). RBM24-mediated upregulation of the maternal allele could provide additive benefit on top of gene replacement.

Computational results — RBM24 motif scan v2 (2026-04-17)

Scripts: ~/STRC/rbm24_scan.py (v1, discarded) → ~/STRC/rbm24_scan_v2.py (correct motifs from ENCODE RBNS ENCSR742AEU).

v1 used UCUUC/CCUCC from literature assumptions. ENCODE RNA Bind-n-Seq shows the actual RBM24 preference is (TG)n repeats (GTGTGT enr=5.05, TGTGTG enr=4.57) and AU-rich GCTCTTC (enr=5.08). v1 results discarded.

v2 results — 5 TG-repeat clusters + 2 AU-rich hits

Exon 4:  pos 1260  — TGTGTG cluster       (outside mini-STRC, N-terminal)
Exon 4:  pos 1529  — TGTGTG cluster       (outside mini-STRC)
Exon 4:  pos 1587  — TGTGTG cluster       (outside mini-STRC)
Exon 11: pos 3014  — GTGTGT cluster       (inside mini-STRC window)
Exon 14: pos 3432  — 4 overlapping hits   (inside mini-STRC) ← STRONGEST
Exon 7:  pos 2469  — GCTCTTC AU-rich      (inside mini-STRC)
Exon 15: pos 3523  — GCTCTTC AU-rich      (inside mini-STRC)

Exon 14 is the primary finding:

• 66 nt (small exon — classic alternative splicing target)
• 4 overlapping TG hits at pos 3432/3433 — densest cluster in the full transcript
• Encodes residues ~1104–1126 — inside the mini-STRC window (700–1775)
• 66 nt ÷ 3 = 22 codons → skipping is in-frame (domain deletion, not frameshift)

Testable prediction: RBM24 loss → exon 14 skipping in STRC mRNA → deletion of amino acids ~1104–1126 → non-functional STRC. Verifiable by RT-PCR in RBM24-null cochlear cells or organoids.

4. mRNA-LNP delivering RBM24 (exon-4 isoform) — instead of AAV, deliver synthetic mRNA encoding the exon-4-containing RBM24 isoform via lipid nanoparticles. No size constraint, no genomic integration, re-dosable. RBM24 protein transiently expressed → corrects STRC splicing. See STRC mRNA Therapy Hypothesis for full analysis.

Critical unknown

RBM24 regulates many targets beyond STRC. Off-target effects of boosting RBM24 in OHCs are unknown. If RBM24 is a global splicing regulator, upregulating it may disrupt other essential transcripts.

Computational next steps

1. RBM24 binding motif scan — Done. Primary target: exon 14 (4-hit TG cluster, residues ~1104–1126). Also exon 11 and AU-rich sites in exons 7/15. v1 discarded; v2 uses ENCODE RBNS data.
2. Validate with eCLIP data — pull RBM24 eCLIP from ENCODE, check if exon 7 (pos 2462/2471) and exon 11 (pos 3077/3093) are confirmed binding sites experimentally.
3. Predicted effect of RBM24 loss on STRC isoforms — use rMATS or MAJIQ to model differential splicing. Does it cause skipping of exon 7 or 11? What protein results?
4. Expression level modeling — if RBM24 upregulation increases STRC 2×, does that cross the HTC coupling threshold? Feed into STRC Stereocilia Bundle Mechanics Model.
5. Off-target scan — pull RBM24 eCLIP, intersect with OHC-expressed transcriptome. Estimate collateral disruption risk from RBM24 overexpression.

Open questions

• Fold-increase in STRC expression when RBM24 is activated?
• RBM24 expression in mature OHCs (adult), or only during development?
• Can RBM24 overexpression restore STRC in a hypomorphic mouse model?
• What is the RBM24 binding-site architecture on STRC pre-mRNA?

Connections

• [applies] STRC Hearing Loss — alternative therapeutic mechanism (RNA-level, not gene delivery)
• [source] 2026-04-17-sun-rbm24-strc-splicing — RBM24 → STRC splicing axis
• [see-also] STRC Mini-STRC Single-Vector Hypothesis — parallel track. Mini-STRC replaces the gene; RBM24 boosts endogenous. Could be additive.
• [see-also] STRC Stereocilia Bundle Mechanics Model — needed to determine if RBM24 boost crosses HTC coupling threshold
• [see-also] Prime Editing for STRC — PE corrects DNA mutation; RBM24 acts at RNA level. Parallel for compound heterozygotes.
• [see-also] STRC AAV Vector Design — RBM24 has a much smaller CDS (~1.5 kb) — fits trivially in AAV
• [see-also] STRC mRNA Therapy Hypothesis — mRNA-LNP delivery of RBM24 exon-4 isoform; same upstream target, different vehicle
• [about] Misha
• Misha-Hearing-10-Year-Plan

Experimental validation — SD03 rMATS analysis (Sun et al. 2026)

Data: ~/Downloads/Alternative splicing of Rbm24 gene/pnas.2531564123.sd03.xlsx Full results: models/rbm24_sd03_splicing_analysis.json

STRC confirmed as direct RBM24 splicing target — 4 significant exon-skipping events (all Down):

Mouse exon (chr2, mm10)	Exon size	IncLevel ctrl	IncLevel KO	Delta	FDR
121203399–121203567	168 nt (56 aa)	94%	40%	-0.542	2.1e-5
121198816–121198894	78 nt (26 aa)	78%	35%	-0.437	3.7e-3
121206192–121206262	70 nt	100%	61%	-0.387	6.3e-3
121198816–121199013	197 nt	88%	63%	-0.263	3.1e-2

Interpretation: when RBM24 exon-4 isoform is absent, STRC loses 40–54% exon inclusion at the strongest site. This is not a minor shift — at 40% delta, roughly half of STRC transcripts are mis-spliced. Non-canonical isoforms likely encode truncated/non-functional stereocilin.

Off-target scope: 469 unique genes, 515 significant SE events across the cochlear transcriptome. RBM24 is a broad splicing regulator — overexpression strategy must account for collateral effects. Off-target analysis needed.