Molecular Vision 2025; 31:175-188
<http://www.molvis.org/molvis/v31/175>
Received 07 December 2024 |
Accepted 14 May 2025 |
Published 16 May 2025
Cheryl Y. Gregory-Evans,1 Aaron W. Joe,2 Kevin Gregory-Evans1
1Department of Ophthalmology and Visual Sciences, University of British Columbia, Vancouver, Canada; 2Kelowna General Hospital, British Columbia, Canada
Correspondence to: Cheryl Gregory-Evans, Department of Ophthalmology and Visual Sciences, University of British Columbia, 2550 Willow Street, Vancouver, BC, V5Z 3N9, Canada; Phone: 1-778-968-4349; FAX: 1-604-675-3010; email: cge30@mail.ubc.ca
Purpose: Members of a multigenerational Canadian family presented to an inherited retinal degeneration (IRD) clinic with retinitis pigmentosa (RP) and sensorineural hearing loss, reminiscent of an Usher syndrome phenotype. Biallelic disease-causing variants in the known Usher syndrome genes were not identified. Therefore, we enrolled further family members in this study and examined whether other IRD gene variants could explain the phenotype in the family.
Methods: Family members underwent a comprehensive ophthalmic examination, including best-corrected visual acuity, direct and indirect ophthalmoscopy, fundus photography, visual field testing, spectral-domain optical coherence tomography, audiological examination, and genetic testing. Some patients also had autofluorescence imaging. Loss-of-function testing was initiated by antisense morpholino knockdown of tubb4b in zebrafish.
Results: Multimodal clinical testing in affected patients revealed an autosomal dominant late-onset presentation of RP associated with progressive, bilateral sensorineural hearing loss that occurred in the second to third decades of life with no vestibular involvement. Panel-based genetic testing revealed a heterozygous c.1168C>T, p.Arg390Trp variant in the beta-tubulin 4B gene (TUBB4B) only in affected family members. Based on in silico analysis, segregation analysis through the family, and literature evaluation, this variant is likely to be the disease-causing variant inherited in an autosomal dominant manner. We searched our local database of ~1,000 patients with IRD, and no other TUBB4B variants were identified, confirming this is a rare disease variant. Knockdown of tubb4b in zebrafish revealed cone and rod photoreceptor abnormalities in the retina and hydrocephalus in the developing brain, resulting in early larval lethality.
Conclusions: For the first time, we describe a multigenerational family with a TUBB4B gene variant p.(Arg390Trp) segregating with deaf-blindness, establishing autosomal dominant inheritance. This further confirms that the Arg390 codon is a mutation hotspot. We also expand the range of phenotypes seen with the p.(Arg390Trp) TUBB4B gene variant to include typical RP as well as a milder, pericentral RP. Furthermore, our studies suggest there is conservation of TUBB4B ciliary function between zebrafish and humans, making zebrafish a better model system for studying vision loss than the mouse model.
Varying degrees of vision loss associated with hearing deficits occur in a highly heterogeneous group of conditions caused by genetic defects, infections, and autoimmune diseases [1-3]. Genetic causes often present from birth to early adulthood in conditions that affect primary cilia, including Usher [4], Bardet-Biedl [5], CHARGE [6], Waardenburg [7], Stickler [8], and Alstrom [9] syndromes. The most commonly seen is Usher syndrome, with a prevalence of between 4 and 17 cases per 100,000 individuals, and accounts for about 50% of all hereditary deaf-blindness cases [10,11], but in pediatric populations, the prevalence is relatively higher at ~1:6,000 [12]. Usher syndrome is characterized by the dual phenotype of sensorineural hearing loss and progressive retinitis pigmentosa (RP), with or without vestibular defects [11,13]. To date, Usher syndrome is known to be caused by defects in 13 genes (RetNet) and is usually inherited in an autosomal recessive manner, although a rare form of the phenotype occurs due to digenic inheritance of mutations in PDZD7 and MYO7A [14].
Four types of Usher syndrome differ based on age of onset, severity of hearing loss, and presence of vestibular involvement (see OMIM 618144). Usher type 1 is the most severe form of the disease, characterized by profound sensorineural hearing loss at birth, abnormal vestibular function, and prepubertal onset of RP [4]. Night blindness and loss of visual fields rapidly progress to registrable blindness by the third decade of life [15]. Usher type 2 (USH2) is the most common subtype, with a congenital mild-to-moderate sensorineural hearing loss, normal vestibular function, and RP that is usually diagnosed in the second decade of life [16]. Usher type 3 accounts for 2% to 4% of cases displaying postlingual sensorineural hearing loss that can progress to profound deafness [17,18]. Vestibular dysfunction occurs in about 50% of patients with Usher type 3 [19], and the onset of RP is usually after the age of 20 years [20]. Usher type 4 represents a group of “atypical” Usher syndrome cases characterized by a late-onset moderate-to-severe sensorineural hearing loss and RP or cone-rod dystrophy with no vestibular involvement [21-24].
Here we report on a family presenting to an inherited retinal dystrophy clinic that appeared to have a form of Usher syndrome that segregated in an autosomal dominant manner. We characterized the phenotype, identified the genetic cause, and modeled the phenotype in zebrafish.
This study was approved by the clinical research ethics board of the University of British Columbia and adhered to the tenets of the Declaration of Helsinki. A four-generation Caucasian Canadian family, originally of Scottish descent, was examined by retinal specialists in an inherited retinal dystrophy (IRD) clinic at the Eye Care Centre in Vancouver, Canada. Informed consent was obtained from all participants included in the study.
Each family member had a comprehensive ocular examination, including best-corrected visual acuity (BCVA), direct and indirect ophthalmoscopy, and slit-lamp biomicroscopy. In addition, color fundus photography (wide-field camera; Optos, Marlborough, MA), visual field testing, autofluorescence, and spectral-domain optical coherence tomography (Heidelberg Engineering, Heidelberg, Germany) were performed. Hearing was tested by standardized pure-tone audiometry (250 to 8,000 Hz) relative to hearing thresholds (normal hearing is ≤19.9 dB hearing loss [HL] in adults). The degree of HL over time was calculated using the pure-tone average (PTA) at four frequencies (0.5, 1, 2, and 4 kHz) in the better ear without amplification [25]. The degree of HL was divided into six categories, including mild (20–34.9 dB HL), moderate (35–49.9 dB HL), moderate to severe (50–64.9 dB HL), severe (65–79.9 dB HL), profound (80.0–94.9 dB HL), and complete loss (≥95 dB HL) [26].
Participant DNA was obtained from buccal swabs, which were screened using panel-based testing of genes associated with retinal dystrophies (Blueprint Genetics, Espoo, Finland). In silico analysis of the variant identified was evaluated with PolyPhen-2, which predicts the impact of a single amino acid substitution on the structure and function of a human protein using physical and comparative considerations; MutationTaster, which analyzes both protein and DNA, thereby assessing single amino acid variations and synonymous, intronic, and deletion variants; and SIFT, which predicts the effects of nonsynonymous single nucleotide variants and frameshifting indels on protein function. The interpretation of pathogenicity was based on criteria established by the American College of Medical Genetics and Genomics [27]. The results were correlated with the participant’s clinical diagnosis, and each participant received genetic counseling regarding their test results.
Research was carried out following protocols compliant with the Canadian Council on Animal Care with the approval of the Animal Care Committee at the University of British Columbia and with the Association for Research in Vision and Ophthalmology statement for the use of animals in vision research. Adult wild-type AB [28] or transgenic Tg[TαCP:GFP] zebrafish [29] strains were maintained at 28.5 °C on a 14-h light/12-h dark cycle in an Aquatic Habitats Aquarium (Pentair Aquatic Eco-Systems, Apopka, FL). Zebrafish were fed brine shrimp (Artemia EG; INVE Aquaculture, Salt Lake City, UT) hatched in the fish facility and flake food for tropical fish (Tetra GMBH, Melle, Germany). In the transgenic Tg[TαCP:EGFP] strain, the cone α-transducin promoter drives enhanced green fluorescent protein (EGFP) expression in all cone photoreceptors. After natural mating, embryos were collected and raised in E3 medium (297.7 mM NaCl, 10.7 mM KCl, 26.1 mM CaCl2, and 24.1 mM MgCl2) containing 1% methylene blue (Sigma-Aldrich, St. Louis, MO) at 28.5 °C.
A splicing-blocking tubb4b-MO1 (5′-AGT AAT TAG ACT CTC ACC TGT GGC T-3′), designed to prevent the messenger RNA (mRNA) splicing of tubb4b exon 2 to exon 3, was synthesized by GeneTools (Philomath, OR). This MO1 blocked both alternate transcripts of tubb4b (tubb4b-201 and tubb4b-202; Zebrafish GRCz11 genome assembly). A control mismatch morpholino oligonucleotide (MO) (tubb4b-MM) was also synthesized (5′-AGT AAT TAG AGT GTC ACG TGT CGC T-3′). A translation-blocking morpholino (tubb4b-MO2) was additionally designed (5′-CCT GTA AAT GCA CGA TCT CCC TCA T-3′) to confirm the phenotype. All tubb4b MOs were diluted in 1X Danieau buffer [58 mM NaCl, 0.7 mM KCl, 0.4 mM MgSO4, 0.6 mM Ca(NO3)2, 5 mM hydroxyethyl piperazineethanesulfonic acid, pH 7.6] containing 0.1% phenol red. The MOs (0.75 ng) were injected into the yolk of one to two cell-stage embryos using a Nanoject II variable injector (Drummond Scientific, Broomall, PA). The morphant embryos were raised at 28.5 °C and observed for morphologic changes using a stereoscopic microscope. To aid image analysis, 0.2 mM phenylthiourea (Sigma-Aldrich, Oakville, ON, Canada) was added to the embryos at 10 h postfertilization (hpf) to inhibit pigment formation. Embryos were fixed in 4% paraformaldehyde (PFA) at specific developmental time points, and images were captured using a Zeiss Axioplan 2 microscope with D63ZNC DVC camera (Carl Zeiss, Toronto, ON, Canada).
We collected 20 wild-type or morphant embryos from the tubb4b-MO2 knockdown at 3 days postfertilization (dpf) for western blotting by initially snap freezing them in liquid nitrogen before
being homogenized by sonication in lysis buffer (10 m
Total RNA was extracted from euthanized zebrafish heads using the Aurum Total RNA Mini Kit (Bio-Rad) according to the manufacturer’s instructions. Using 0.5 µg total RNA, the reverse-transcription reaction was performed in a 20-µl reaction volume containing 4 mM deoxynucleotide triphosphate mix, 1X random primers, 1X RT buffer, 20 U RNase inhibitor, and 50 U RT enzyme (Applied Biosystems High Capacity cDNA Reverse Transcription Kit; ThermoFisher Scientific, Waltham, MA). The yield of complementary DNA product was evaluated by Nanodrop spectrophotometry (ThermoFisher Scientific). We used 5 ng complementary DNA as a template for amplification using the QuantiTect Reverse Transcription Kit (Qiagen, Toronto, ON, Canada). Primers for tubb4b (330 bp product) were forward primer 5′-CAA CAA TGA GGG AGA TCG TG-3′, reverse primer 5′-GTT CCA CGT GCC GTG CTG GT-3′; primers for rhodopsin (405 bp product) were forward primer 5′-AGT CCT GCC CAG ACA TCT AG-3′, reverse primer 5′-CGA CCA TAG CCC CAT CTC AC-3′; primers for tubb6 (268 bp product) were forward 5′-AGT GTT CGT TCT GGA GCG TT-3′, reverse 5′-ATG ATG CGG TCC GGG TAT TC-3′; and primers for gapdh (190 bp product) were forward primer 5′-CAC CAG GGC TGC TTT TAA C-3′, reverse primer 5′-ATC TCG CTC CTG GAA GAT-3′. Thermocycling parameters were as follows: 2 min at 50 °C, 20 s at 95 °C, 40 cycles of 1 s at 95 °C, plus 20 s at 58 °C. ImageJ software (National Institutes of Health, Bethesda, MD) was used to quantify band intensities relative to gapdh.
Four affected members and four unaffected members of a four-generation Canadian family were examined at an IRD specialist clinic with a phenotype of hearing loss and RP. This deaf-blind syndrome segregated in an autosomal dominant manner (Figure 1). Clinical data for this family are presented in Table 1. The male proband (III-4) presented in 2012 at 50 years of age, reporting poor vision since age 23. He also reported hearing loss from 9 years of age. His BCVA at age 50 had been 20/20 right and 20/30 left. Anterior segment examination was unremarkable. Dilated fundus examination revealed mid-peripheral chorioretinal atrophy with bone spicule retinal degeneration extending into the far periphery (Figure 2). Goldmann visual field testing from 2009 showed extensive mid-peripheral field loss with preservation of the central 10 degrees in each eye and some far peripheral field retention (Figure 3). Further testing at 55 years old and 62 years old revealed progression of bone spicule retinal degeneration (Figure 2) and further loss of far peripheral visual field (Figure 3).
The proband’s older brother (III-1) presented at 60 years of age. He reported night blindness from the age of 17 years and deafness from early childhood. His visual acuity at age 60 was 20/40 right and 20/50 left. A dilated fundus examination showed optic nerve pallor in both eyes with extensive mid-peripheral chorioretinal atrophy (Figure 4). Visual field testing at 61 years of age showed near-mid-peripheral loss confined between 5 and 30 degrees, consistent with a diagnosis of pericentral retinitis pigmentosa (Figure 3). Field loss was less severe than in his younger brother.
The proband’s younger sister (III-6) presented in 2023 at 54 years of age and reported vision loss from 32 years of age and hearing loss from 37 years old. BCVA was 20/70 right and finger counting only left. Retinal imaging showing extensive chorioretinal atrophy with visual fields restricted to just 2 degrees or less.
The proband’s son (IV-4) presented in 2021 at 24 years of age with no visual symptoms but recent mild hearing loss. BCVA was 20/20 right and left. Clinical examination revealed minor bone spicule pigmentary changes in the left eye and retinal pigment epithelial anomalies in the mid-periphery of both eyes. Goldmann visual fields were normal.
Both first and second generations of the family were reported by the proband as being deaf and blind but were not examined in the clinic. The proband reported poor hearing since he was 11 years old, and this was not corrected during his school years. He started using hearing aids at age 22. An auditory test at age 28 showed bilateral mild-to-moderate mid-frequency sensorineural hearing loss with a PTA (at 0.5, 1, 2, and 4 K frequencies) of 42.5 dB HL (Figure 5A). The progression to moderately severe was recorded at age 40 (PTA 56.6 dB HL). In the most recent audiometry test at age 60, he had mild-to-moderately severe hearing loss with a PTA of 57.5 dB HL (Figure 5A). His son (IV-4) at age 9 had a normal audiometry test, but at age 16, he was found to have a mild bilateral sensorineural hearing loss in the 1,000 to 4,000 Hz range (PTA 26.25 dB HL; Figure 5B). At age 22, this had progressed to moderate hearing loss (PTA 43.7 dB HL; Figure 5B), and thus he started wearing hearing aids. The elder brother of the proband (III-1) had hearing loss since childhood, and his most recent audiometry test at age 57 showed he had bilateral moderately severe to severe sensorineural hearing loss with a PTA of 67.5 dB HL (Figure 5C). The sister of the proband (III-6) noticed hearing loss in the third decade of life. Her most recent audiometry test was at age 51, when she exhibited bilateral mild-to-moderate sensorineural hearing loss, with a PTA of 45.0 dB HL (Figure 5D). She started wearing hearing aids at age 51. Family members III-3, IV-1, IV-2, and IV-3 did not report any hearing loss.
Since the clinical phenotype resembled that of Usher syndrome, we expected to see variants in the known Usher syndrome genes. However, upon panel-based genetic testing in 2019 (266 genes), the proband had three variants of uncertain significance: ADAMTS18 [c.570C>G, p.(Asn190Lys)], USH2A [c.12611C>T, p.(Thr4204Met)], and WFS1 [c.2494C>T, P.(Arg832Cys)]. ADAMTS18 and USH2A are inherited in a recessive manner, so they are unlikely to explain the phenotype in the patient. In addition, in the Leidon Open Variation Database for USH2A, the p.(Thr4204Met) USH2A variant is reported as benign. WFS1 variants cause Wolfram syndrome (optic atrophy, sensorineural hearing loss, and dysarthria) and can be inherited in recessive and dominant forms. However, the specific WFS1 variant we found is relatively common in the gnomAD reference database, and the patient did not have optic atrophy, so it seems unlikely to be the cause of the phenotype. Another member of the family (III-1) was tested in 2021 when the gene panel was larger (285 genes), and a single missense variant in the TUBB4B gene (c.1168C>T, p.(Arg390Trp) was identified, which was previously reported as a mutation hotspot [30]. In silico modeling for this variant with Polyphen-2 predicted it was probably damaging; with SIFT modeling, it was predicted to be deleterious; with MutationTaster modeling, the prediction was disease-causing. Missense variants in TUBB4B (Table 2) had been recently reported in association with dominantly inherited Leber congenital amaurosis with early-onset deafness [31]. We therefore tested the rest of the family and retested the proband (as TUBB4B was not on the genetic testing panel at first testing). Indeed, we found all affected family members had the same TUBB4B variant, and none of the unaffected members carried the variant. Importantly, family member IV-4 did not have the WFS1 variant that his father had, ruling out WFS1 as a cause of the dominant disease in the family.
We used morpholino gene knockdown technology to provide functional evidence that tubb4b is important to ocular development [32]. Injection of a mismatch control morpholino did not affect wild-type at 72 hpf, but injection of either splice-blocking or translation-blocking tubb4b morpholinos resulted in morphants displaying cardiac edema and hydrocephalus (Figure 6). Immunohistological analysis of the retina revealed a loss of cone photoreceptors, as demonstrated by a reduction in cone green fluorescent protein (GFP) expression in the TαCP:EGFP transgenic retina for both morpholinos. Confirmation of the effectiveness of splice-blocking MO1 was determined by reverse transcription (RT)-PCR since this would result in loss of tubb4b exon 2 in the mRNA transcript (Figure 7A,B). Confirmation of the effectiveness of translation-blocking morpholino MO2 was determined by tubb4b western blotting, where tubb4b protein was not detected (Figure 7C). To test for loss of rod photoreceptors in embryos injected with translation-blocking MO2, we used semiquantitative RT-PCR to examine the relative number of rhodopsin transcripts and found an 82% ± 7% (standard error of the mean, n = 3) reduction, reflecting a lower number of rods present in the retina (Figure 7D). In a recently reported mouse Tubb4b knockdown experiment, Tubb6b was upregulated as a compensatory mechanism for the loss of Tubb4b [33]. Using semiquantitative RT-PCR normalizing gene expression to GAPDH, we found that expression of tubb6 was unchanged in the tubb4b morphants (Figure 7D).
In our study, we identified a four-generation family segregating a p.(Arg390Trp) TUBB4B variant with RP and late-onset, progressive sensorineural hearing loss inherited in an autosomal dominant manner. To our knowledge, this is the first reported family with this phenotype. There is recent literature supporting TUBB4B variants in association with retinopathy segregating with hearing loss at two mutation hotspots [30]. Two ultra-rare variants in the TUBB4B gene were initially reported to segregate with Leber congenital amaurosis with early-onset deafness: p.(Arg391His) and p.(Arg391Cys) [31,34]. A de novo variant p.(Gln11Arg) was also identified in an American female child with hearing loss, Faconi syndrome, rickets, and microphthalmia but without retinal abnormalities [35]. Four unrelated de novo cases had a p.(Pro358Ser) TUBB4B variant that was associated with Leber congenital amaurosis, sensorineural hearing loss, and primary ciliary dyskinesia [36].
Also recently, Bodenbender and colleagues described three variants [30]: a p.(Arg390Gln) variant segregated in a mother and son with late-onset night blindness and symptomless-to-moderate hearing loss; a de novo p.(Tyr310His) variant was found in a female infant with congenital profound hearing loss and reduced electroretinographic responses but without a specific retinal diagnosis. Most importantly, Bodenbender and colleagues reported a p.(Arg390Trp) variant in a single 36-year-old patient who reported glare sensitivity only and hearing loss. On further analysis, he was found to have a reduced electroretinographic response and thinning of the outer retina on optical coherence tomography imaging. Segregation of this variant was not available.
To our knowledge, the study is the first to report the TUBB4B variant p.(Arg390Trp) segregating with disease in an extended pedigree, establishing autosomal dominant inheritance. A comparison of retinal phenotypes also distinguishes the TUBB4B variant p.(Arg390Trp) phenotype seen in our study in comparison to the TUBB4B variant p.(Arg390Trp) phenotype reported by Bodenbender and colleagues [30]. In that simplex case, the retinal phenotype was described as a pericentral RP, whereas proband III-4 in our study is clearly better described as having a more extensive, typical RP (Figure 2 and Figure 3). Participant III-1 could be described as having a case of pericentral RP, which is described as milder and more slowly progressive than typical RP (Figure 3 and Figure 4). We conclude that unlike the report in Bodenbender and colleagues [30], we cannot confirm a solely mild retinal phenotype to the TUBB4B variant p.(Arg390Trp) but more accurately portray it as variable, with pericentral or typical retinitis pigmentosa subtypes possible.
It can also be concluded from the literature to date that six TUBB4B variants targeting four codons are associated with deaf-blindness. Variants at either codon Arg390 or Arg391 are restricted to phenotypes with hearing loss and retinal disease, whereas codons Gly11 and Pro358 are associated with deaf-blindness and a broader range of systemic ciliopathies.
To understand how variants in TUBB4B cause retinal and hearing loss phenotypes, we knocked down the tubb4b gene in zebrafish and found loss of both rod and cone photoreceptors, as well as hydrocephalus, which caused lethality by 5 dpf. In the developing otic vesicle, we did not observe structural abnormalities in the otoliths (ear stones) at 72 hpf, which are connected to the kinocilia of hair cells and are necessary for detecting sound and sensing linear accelerations in zebrafish [37]. However, the developing inner ear does not reach full maturity until 20 dpf [38], and thus we cannot be certain that there are no hearing defects in the tubb4b morphant larvae. Since the knockdown is lethal, we were not able to take advantage of detecting auditory defects at 5 dpf by analyzing startle reflex responses [39]. Future studies creating zebrafish tubb4b knock-in mutants may allow long-term studies to resolve whether the mutants have hearing loss and for testing possible therapeutic approaches.
In two recent studies, targeted deletion of the Tubb4b gene in mice resulted in hydrocephaly and perinatal lethality [36,40] similar to our zebrafish data. Generation of a mouse line carrying a mutation equivalent to a known human mutation (Tubb4bR391H/+) resulted in mice that were infertile due to a spermatogenesis defect, but they did not develop any retinal degeneration [37]. The mice had profound deafness due to motility defects in epithelial cells of the middle ear and decreased microtubules in pillar support cells in the inner ear. It was further demonstrated that a compensatory upregulation of Tubb6 mRNA and protein in the knockout mice could be why there was an absence of a retinal phenotype in mouse mutants [33]. We also tested tubb6 expression in the zebrafish morphants and did not find any upregulation of tubb6 mRNA, suggesting there is no compensatory mechanism in zebrafish. Thus, although the Tubb4b mouse models are excellent for studying hearing loss, they are not useful for studying vision loss. One possibility may be to generate a Tubb4b mutant on a Tubb6 knockout background to elicit an eye phenotype. A similar disconnect in phenotype is seen with mouse models of Usher syndrome type 1B due to abnormalities in the Myo7a gene, in which hearing loss is present with only subtle retinal phenotypes [41]. This has been explained by a difference in the structure of mouse and human photoreceptors, where calyceal processes expressing MYO7A are absent in mice [42]. Further studies in human tissue analyzing expression of TUBB4B in calyceal processes may explain the lack of phenotype in mice.
In summary, we have identified a TUBB4B disease-causing variant in a four-generation family segregating deafness and RP, which expands the clinical phenotype associated with TUBB4B gene variants. To our knowledge, this is the first time a TUBB4B variant has been reported as segregating with disease in an extended family, establishing autosomal dominant inheritance. We also expand the range of phenotype seen with the typical RP phenotype in addition to the pericentral phenotype previously documented. This phenotype could represent a dominant form of Usher syndrome type 3, characterized by postlingual sensorineural hearing loss and RP without vestibular involvement. Due to the high level of gene conservation, loss of function of Tubb4b/tubb4b leads to early lethality in both mice and zebrafish and suggests that homozygous variants in TUBB4B in humans might also be embryonic lethal. Since the tubb4b zebrafish model demonstrates an eye phenotype unlike mice, this could become the best model to test potential therapeutic options for the retinal phenotype.
This work was supported by the Fighting Blindness Canada Patient Registry (grant number F22-04266_20240521). Declaration of interest statement. The authors state they have no conflicts of interest to declare.