DFNA68: HOMER2-Related Autosomal Dominant Nonsyndromic Hearing Loss — Comprehensive Research Report

Prepared: 2026-09-19. Evidence base: the six primary DFNA68 family reports (2015–2026) plus the supporting mechanistic literature; all quoted snippets below are verified exact quotes from the cited abstracts/records.

1. Disease Information

Overview. DFNA68 is a bilateral, postlingual, progressive, nonsyndromic sensorineural hearing loss (SNHL) caused by heterozygous variants in HOMER2, which encodes a scaffolding protein of the Homer family enriched at the tips of cochlear hair cell stereocilia. Hearing is normal at birth and through early childhood; loss typically begins in the first or second decade at the high frequencies, producing a down-sloping audiogram that flattens as lower frequencies are drawn in, and deteriorates at a measured 1.2–1.6 dB/year until severe or profound. Fewer than ten unrelated families have been reported worldwide since the gene was identified in 2015 (PMID:25816005), making this one of the rarest characterized DFNA loci.

Identifiers:

Synonyms: DFNA68; deafness, autosomal dominant 68; autosomal dominant nonsyndromic deafness 68; HOMER2-related hearing loss.

Data provenance: All clinical information is aggregated from published family reports (pedigree/case-series literature), not EHR-derived. The entire clinical description rests on approximately six pedigrees plus one isolated patient.

Key primary literature (PMIDs): 25816005 (founding study, 2015); 30047143 (second family, Chinese, 2018); 33809266 (Spanish family, 2021); 37173411 (nonstop variant, 2023); 40515814 (frameshift extension + cochlear implant outcome, 2025); 41351289 (Chinese four-generation family, 2026).

2. Etiology

Causal factor: Purely genetic — heterozygous germline variants in HOMER2 (15q24.3; OMIM places it at 15q25). No environmental, infectious, or lifestyle cause or modifier has been reported in any DFNA68 family.

Critically, this is not a haploinsufficiency disease. Three independent lines of evidence converge: - Mice heterozygous for a Homer2 null hear normally; only the homozygous null is deaf (PMID:25816005; restated in PMID:33809266: "mice heterozygous for the targeted deletion of exon 3 in Homer2 (Homer−/+) displayed normal hearing levels"). - HOMER2 is tolerant of loss-of-function variation in population data: "Low pLI score (0.01) of HOMER2 suggests that alteration of HOMER2 exerts a pathogenic effect via a mechanism other than loss-of-function" (PMID:40515814). - The founding allele was shown experimentally to act as a dominant negative: zebrafish mutant-mRNA injection produced an ear phenotype while knockdown of the endogenous gene did not (PMID:25816005).

Risk factors: Being an obligate carrier in an affected family (50% transmission risk). No susceptibility loci, modifier genes, protective variants, or gene–environment interactions have been reported. No known environmental risk or protective factors specific to this disorder (standard noise/ototoxin avoidance advice for any progressive SNHL applies on general principles only — not disease-specific evidence).

3. Phenotypes

The phenotype is essentially monosymptomatic. Suggested HPO annotations with frequencies (frequencies reflect the small reported family set, not measured cohort proportions):

Key supporting quotes: "Hearing impairment had a typical onset in the first decade of life in the high frequencies, with significant subsequent progression of hearing loss over all frequencies" and "The resulting annual threshold deterioration (ATD) was 1.2 to 1.6 dB per year" (PMID:25816005).

Substantive negatives (actively tested, not just unreported): - No vestibular dysfunction: "Tandem gait and Romberg tests did not reveal any symptoms of vestibular dysfunction in all family members" (PMID:41351289). - No syndromic features: "Clinical examination was negative for any findings consistent with syndromic hearing loss and also ruled out autoimmune phenotypes" (PMID:25816005). Intelligence and speech normal (PMID:41351289).

Quality of life: Documented at the severe end: one patient reached 8% bilateral speech discrimination ten years after significant progression began, requiring cochlear implantation (PMID:40515814). No formal QoL instrument (EQ-5D/SF-36) data exist for this disorder.

4. Genetic/Molecular Information

Gene: HOMER2 (hgnc:17513; OMIM 604799), chromosome 15q24.3, nine exons, two transcripts encoding 343-residue (NM_004839) and 354-residue (NM_199330) isoforms. Nomenclature trap:* both numbering systems appear in the literature for the same variants — the founding missense is p.Arg185Pro (isoform 1) and p.Arg196Pro (isoform 2); these are one allele, not two.

Protein architecture: N-terminal EVH1 domain (binds proline-rich motifs) + C-terminal coiled-coil (CC) domain that (a) drives tetramerization and (b) carries the CDC42-binding domain (CBD, ~residues 191–283; mapped in neurons, PMID:19309525).

Reported pathogenic variants — three mechanistic classes (all germline, heterozygous; ~6–7 variants in fewer than ten families; the two published tallies — "five variants" in PMID:41351289 vs "seventh variant" in PMID:40515814 — disagree by one and were never reconciled):

1. Coiled-coil missense (dominant negative): c.554G>C, p.Arg185Pro/p.Arg196Pro (European kindred, PMID:25816005). "These results show that HOMER2 plays an essential role in the normal development and/or maintenance of hair cells in the zebrafish inner ear and that the p.Arg185Pro mutation has a dominant-negative effect on this process."
2. Frameshift truncations within the CDC42-binding domain (dominant negative / possible GoF): c.840_841insC (Chinese, PMID:30047143) and a five-base deletion producing p.Pro278Alafs*10 (Spanish, PMID:33809266). The truncated product shows "decreased protein stability, reduced ability to multimerize, and altered distribution pattern in transfected cells." The Spanish authors explicitly leave open whether the mechanism is "a gain-of-function of the Ct aberrant tail or ... a dominant-negative mechanism."
3. C-terminal extension variants (argued gain-of-function): nonstop c.1064A>G converting the stop codon to tryptophan (+10 residues; transcripts escape nonstop decay, PMID:37173411); c.1033delC p.Arg345Glufs*64 (+~64 residues; PMID:40515814 — "Our findings reveal a solid pathogenic gain-of-function effect related to the long extension to the C-terminal of HOMER2"); c.1023_1029del p.Asp342ArgfsTer54 (PMID:41351289). AlphaFold2 predicts the longest extension disrupts the distant EVH1 domain and its Cdc42 interface while preserving multimerization (computational, not measured).

Population frequency: Disease alleles are absent/ultra-rare in population databases; the gene itself tolerates LoF (pLI 0.01), which is the population-genetic argument against haploinsufficiency.

Modifiers, epigenetics, chromosomal abnormalities: None reported. No recessive HOMER2 patient has been reported, despite the founding authors' prediction: "The recessive phenotype exhibited by null alleles of Homer2 makes it a strong candidate for autosomal recessive hearing loss due to loss of function in humans as well" (PMID:25816005) — still open a decade later.

5. Environmental Information

Not applicable. No environmental, lifestyle, or infectious contributor has been reported in any DFNA68 family. This is a fully penetrant-appearing Mendelian mechanosensory disorder as described (though formal penetrance quantification is lacking — see §9).

6. Mechanism / Pathophysiology

Ordered causal chain (initiating lesion → clinic):

1. A heterozygous HOMER2 variant (missense in the coiled coil, frameshift truncation in the CDC42-binding domain, or C-terminal extension) produces a mutant scaffold protein that is expressed, not eliminated (NMD escape shown/inferred for the extension and truncation classes).
2. The mutant protein interferes with wild-type HOMER2 — impairing tetramer assembly for the missense/truncation classes (demonstrated in vitro for the truncation: reduced multimerization, instability, mislocalization), or adding an aberrant C-terminal tail (extension class; gain-of-function argued from a designed zebrafish truncation-vs-extension comparison). This step is why the disease is dominant despite a phenotype-free null heterozygote.
3. Defective scaffold hubs lead to disrupted HOMER2–CDC42 coupling at the stereocilia (the CBD sits inside the coiled coil; for the extension class, the route is a predicted EVH1 conformational change) — inferred, not demonstrated: no binding assay exists for any disease allele, and no HOMER2–CDC42 complex has been shown in a cochlear hair cell.
4. Branch (alternative/parallel hypothesis): impaired scaffolding leads to dysregulated hair-cell calcium handling via Homer's regulation of PMCA pumps and TRPC channels — hypothetical: the one real measurement is in mouse parotid gland (PMID:25049230), PMCA2 being "the only system for clearance of Ca2+ from hair cell stereocilia" is what makes it attractive (PMID:25816005), and the established Homer2 TRPC partner (TRPC1) is not the TRPC3/TRPC6 hair cells use. The founding paper named both hypotheses and "further functional studies are needed" — nothing since has closed them.
5. Loss of CDC42-tuned actin turnover results in failure of stereocilia maintenance — bundles form normally, then deteriorate. Inferred by analogy: hair-cell-specific Cdc42 knockout mice show exactly this: "Hair cells of Atoh1-Cre;Cdc42(flox/flox) mice developed normally but progressively degenerated after maturation, resulting in progressive hearing loss particularly at high frequencies", beginning as "stereocilia fusion and depletion" (PMID:24610943). No stereocilia morphology has been reported for any HOMER2 patient or model.
6. Bundle failure results in hair-cell functional failure without hair-cell death — the mechanistically distinctive claim, and the step the models actually establish: the homozygous null mouse is profoundly deaf with a full complement of inner and outer hair cells at 8 weeks ("No differences in IHCs and OHCs were observed in any animals regardless of genotype", PMID:25816005), and zebrafish injected with disease alleles show "severely impaired ... capacity of neuromast hair cells to uptake FM1-43" with "no significant differences in hair cell numbers" (PMID:40515814).
7. Failed mechanotransduction leads to progressive bilateral sensorineural hearing loss, postlingual, high-frequency-first, at 1.2–1.6 dB/year — the clinical endpoint.

Ontology suggestions: GO:0097110 scaffold protein binding; GO:0051289 protein homotetramerization; GO:0031267 small GTPase binding; GO:0030036 actin cytoskeleton organization; GO:0060088 auditory receptor cell stereocilium organization; GO:0051480 regulation of cytosolic calcium ion concentration; GO:0050910 detection of mechanical stimulus involved in sensory perception of sound; GO:0007605 sensory perception of sound; cellular components GO:0032420 stereocilium, GO:0032426 stereocilium tip; cell types CL:0000589 cochlear inner hair cell, CL:0000601 cochlear outer hair cell.

No molecular profiling (transcriptomics, proteomics, metabolomics, single-cell) has been published for this disorder specifically.

7. Anatomical Structures Affected

• Organ: inner ear / cochlea only (UBERON:0001846 internal ear; UBERON:0001844 cochlea; UBERON:0002227 organ of Corti). No secondary organ involvement in any patient (a zebrafish cardiac signal for the longest extension allele is an open question — the patient carrying it had no cardiac abnormality; see §11/§15).
• Cells: cochlear inner and outer hair cells (CL:0000589, CL:0000601). Mouse Homer2 expression is "particularly enriched in the tips of stereocilia of both IHCs and OHCs" (PMID:25816005).
• Subcellular: stereocilium and stereocilium tip (GO:0032420, GO:0032426); the actin core and apical junctional actin belt (by CDC42-model analogy).
• Lateralization: bilateral, symmetric.

8. Temporal Development

• Presymptomatic phase: normal hearing through infancy; passes newborn hearing screening; carriers identified genetically before audiometrically.
• Onset: postlingual, typically first–second decade; 17–24 years in the most recent Chinese family; one patient's significant progression began in her 50s. Insidious.
• Progression: relentlessly progressive, non-episodic, lifelong; 1.2–1.6 dB/year across frequencies (founding family, threshold-on-age regression). Audiometric configuration is stage-dependent: high-frequency down-sloping early (or, rarely, low-frequency ascending), flattening to all-frequency involvement, ending severe-to-profound.
• No remission is described; no critical intervention window beyond ordinary timing of amplification/implantation.

9. Inheritance and Population

• Inheritance: autosomal dominant (HP:0000006), segregating across up to five generations. Every reported allele is heterozygous; no homozygote or compound heterozygote described.
• Penetrance/expressivity: appears complete in reported pedigrees but is formally unquantified (age-dependent onset means young carriers look unaffected). Expressivity is variable mainly in onset age and audiogram configuration. No anticipation, germline mosaicism, founder effect, or consanguinity role reported.
• Epidemiology: no prevalence or incidence estimate exists; "Hearing loss caused by HOMER2 mutations is an extremely rare disorder" (PMID:33809266). Ultra-rare; countable in families: European, Chinese (×2), Spanish, Sicilian, Italian origins plus one isolated patient. No sex bias apparent in the pedigrees. For context, dominant nonsyndromic hearing loss overall involves ~50 genes at ~67 loci (PMID:33809266, citing the field count).

10. Diagnostics

• Route to diagnosis is molecular, not clinical. There is no biomarker, no biochemical test, and no audiometric pattern specific enough to distinguish DFNA68 from the ~50 other DFNA genes. Every family was diagnosed by comprehensive hearing-loss gene panel or whole-exome sequencing followed by Sanger co-segregation. "Our study also highlights the importance of using NGS-based diagnostic methods to identify mutations in low-prevalence deafness genes like HOMER2" (PMID:33809266). The practical issue is panel content: HOMER2 must be on the ordered panel to be found.
• Audiology: pure-tone audiometry with bone conduction (excludes conductive loss) and serial follow-up — a single audiogram describes a stage, not the disorder; the deterioration rate requires threshold-on-age regression. ABR and speech discrimination scoring at the severe end.
• Imaging: the one temporal-bone CT reported was normal — "CT scans of the proband ruled out middle- and inner-ear malformations" (PMID:41351289). Imaging excludes alternatives rather than making the diagnosis.
• Differential: other DFNA loci (notably other progressive high-frequency forms such as DFNA2/KCNQ4), excluded syndromic mimics (vestibular testing and syndromic examination were normal — see §3).
• Screening: newborn hearing screening does not detect DFNA68 (postlingual onset). Cascade genetic testing of at-risk relatives is the applicable screening mode.
• No GeneReviews chapter exists for DFNA68 or HOMER2 (verified against the full GeneReviews index, snapshot 2026-09-10).

11. Outcome/Prognosis

• Survival: normal; no mortality or systemic morbidity attributable to the disorder.
• Functional trajectory: progressive to severe/profound deafness in an adult who acquired speech normally. Best-documented endpoint: "Ten years after the onset of significant SNHL, SB1190-1923 had developed profound deafness with a speech discrimination score of only 8%" (PMID:40515814).
• Prognostic factors: allele class may modulate onset/severity (the longest extension allele was argued to be more severe than a matched artificial truncation in zebrafish), but human genotype–phenotype correlation is not established across so few families. The measured 1.2–1.6 dB/year rate is the single quantitative prognostic anchor.
• Recovery potential: hearing does not recover; cochlear implantation restores speech recognition (below).
• Open cardiac question: zebrafish overexpressing the longest extension allele showed increased cardiac anomalies ("a possible link to cardiac anomalies", PMID:40515814); no cardiac feature has been described in any DFNA68 patient, so this remains an unconfirmed model-system signal, not a prognosis item.

12. Treatment

No disease-modifying, gene-directed, or pharmacological therapy exists or is in trials; no HOMER2-directed agent and no NCT identifier appears in any primary report. Management is device-based and supportive:

1. Hearing aids (first line while thresholds permit; NCIT note: hearing-aid fitting has no NCIT clinical-action term; NCIT:C183182 is the device concept). The disorder-specific point is commitment to repeated audiological review given measurable annual progression: "A long audiological follow-up is of paramount importance to identify hearing threshold deteriorations early and ensure prompt treatment with hearing aids or cochlear implants" (DFNA class-level review, PMID:37371710).
2. Cochlear implantation (NCIT:C15329 Surgical Procedure + device NCIT:C157820 Cochlear Implant) once amplification fails. DFNA68 has one published genotype-confirmed outcome: a patient in her 60s with 8% speech discrimination implanted, with "speech recognition improves significantly at 3- and 6-months post-implantation across various syllable and sentence conditions" (PMID:40515814). Mechanistically favorable: the lesion is hair-cell-level and non-degenerative (hair cells and, presumably, spiral ganglion neurons preserved), upstream of what the implant stimulates.
3. Genetic counselling and cascade testing (NCIT:C15240 Genetic Counseling): 50% transmission risk; a normal audiogram in a young relative is uninformative (onset can be delayed to the late teens/twenties or beyond); the molecular diagnosis mainly buys a predicted trajectory, since it selects no therapy.

Pharmacogenomics, gene therapy, ASO/RNA therapy: none exists for DFNA68. (The dominant-negative/gain-of-function allele biology would make allele-selective knockdown a rational future strategy — as pursued in other DFNA genes — but no such program has been reported for HOMER2; this is an inference, not a published claim.)

13. Prevention

• Primary prevention: none possible (germline genetic).
• Secondary: cascade genetic testing in known families; serial audiometry in carriers to time amplification. Newborn hearing screening does not capture it.
• Tertiary: timely re-fitting of amplification and timely implant referral to limit auditory deprivation.
• Reproductive options: standard for an AD condition (prenatal/preimplantation testing where a familial variant is known) — general genetic-counselling practice; no DFNA68-specific guideline exists.

14. Other Species / Natural Disease

No naturally occurring HOMER2-associated deafness has been reported in companion animals or wildlife (no OMIA phenotype record identified). Orthologs are conserved across vertebrates — mouse Homer2 (MGI) and zebrafish homer2 (~67% identity to human protein) are the studied orthologs. Not zoonotic; not applicable.

15. Model Organisms

The central model gap: no system carries a single copy of a human DFNA68 allele expressed at its own level in a mammalian cochlea. A heterozygous knock-in allelic series (missense, CBD-frameshift, extension) with ABR/DPOAE follow-up, SEM bundle morphology, and late hair-cell counts is the experiment that would convert the dominant-negative and gain-of-function classifications from inference to measurement — and settle whether hair cells truly survive decades of the lesion.

Key open questions (for knowledge-base gap tracking)

1. Actin vs. calcium: does HOMER2 act in hair cells through CDC42/actin, through PMCA2/TRPC calcium handling, or both? Proposed in 2015, still untested; no binding assay for any disease allele, no HOMER2 partner mapping in cochlear tissue.
2. No model of the human genotype (above).
3. Cardiac signal of the extension allele: zebrafish-only; echocardiography of existing carrier cohorts would resolve it.
4. Recessive HOMER2 hearing loss: predicted from the mouse null in 2015, never observed.

Sources

• OMIM #616707 — Deafness, autosomal dominant 68 · OMIM *604799 — HOMER2
• Azaiez et al. 2015 — HOMER2, a stereociliary scaffolding protein, is essential for normal hearing in humans and mice (PMID:25816005)
• Lu et al. 2018 — WES identified a second pathogenic variant in HOMER2 (PMID:30047143)
• 2021 — Novel truncating mutation in HOMER2, Spanish family (PMID:33809266) (PMC8001007, DOI:10.3390/genes12030411)
• 2023 — HOMER2 nonstop variant, EJHG (PMID:37173411) (Nature/EJHG)
• 2025 — Impact of HOMER2 frameshift extension variant on auditory function and development (PMID:40515814) (PMC12343744)
• 2026 — Novel frameshift c.1023_1029del in HOMER2, J Clin Lab Anal (PMID:41351289)
• Ueyama et al. 2014 — Maintenance of stereocilia by Cdc42 in cochlear hair cells (PMID:24610943)
• DFNA narrative review 2023 (PMID:37371710)
• Orphanet ORPHA:90635 — AD non-syndromic sensorineural deafness type DFNA
• LOVD Disease #06261 (DFNA68) · GeneCards HOMER2

Reference Validation

Checked with linkml-reference-validator 0.2.1.

All extracted references resolved successfully.

Term Validation

Checked with linkml-term-validator 0.4.5, through the ols: adapter.

Terms whose name is worth a second look

The report's name for these is recognisably related to the term's own name without being one of them. A loose paraphrase reads the same way as a citation of the wrong sibling term - and so does a related synonym, which the ontology records precisely because it names something adjacent rather than the same thing - so these are listed rather than judged:

• HP:0008596 (1 mention) - the report calls it "Postlingual onset"; HP calls it Postlingual sensorineural hearing impairment
• HP:0008619 (1 mention) - the report calls it "Bilateral involvement"; HP calls it Bilateral sensorineural hearing impairment, and lists "Bilateral nerve deafness" among its other names
• HP:0001757 (1 mention) - the report calls it "High-frequency-onset, down-sloping loss"; HP calls it High-frequency sensorineural hearing impairment
• HP:0008573 (1 mention) - the report calls it "Low-frequency (ascending) loss at presentation"; HP calls it Low-frequency sensorineural hearing impairment, and lists "Low-frequency sensorineural hearing loss" among its other names

Prefixes with no resolver

Terms carrying these prefixes were not checked either way, because no configured ontology covers them. An unrecognised prefix may name an ontology this run could not reach as easily as one that does not exist, so nothing here is evidence of fabrication: ORPHA.

27 of 28 terms resolved to a current term; the rest could not be looked up either way.