Ultrarare missense and frameshift variants in the TECTA gene may involve tectorial membrane in familial Meniere disease

Background: Meniere’s disease (MD) is an inner ear disease defined by episodes of vertigo associated with sensorineural hearing loss initially affecting low- to medium frequencies, tinnitus, and aural fullness. Familial aggregation has been reported in 9 -10% of MD patients showing, mostly, an autosomal dominant inheritance pattern with incomplete penetrance. However, familial MD is a genetically heterogeneous disorder and other inheritance patterns have been recently proposed, such as recessive and digenic inheritance involving rare variants in OTOG and MYO7A genes, respectively. In this study , a familial MD cohort was recruited to identify new candidate genes.

Exome sequencing was performed in 99 individuals (from 77 families) diagnosed with MD according to the diagnostic criteria define d by the Barany Society . Candidate variants were classified based on the ACMG/AMP guidelines , and their effects were evaluated by protein modeling. Standard audiometric evaluations were retrieved, and a case report was made of each family to assess the genotype-phenotype correlations.

The TECTA gene, which encodes α -tectorin, was highlighted as a candidate for four multicase MD families carrying rare missense heterozygous variants and a short del etion in this gene. Variants in α-tectorin were also found in two additional families with one MD patient and relatives with partial syndromes carrying a missense heterozygous variant and a short deletion . According to the predicted protein model, these variants could affect the stability of α-tectorin.

Several MD families were identified carrying rare variants and deletions in t he TECTA gene, which encodes one of the main proteins of the tectorial membrane (TM). The TM is an extracellular matr ix localized over the sensory epithelium mediating the mechanical stimulation of cochlear hair cells, a critical role in the process of hearing. Modifications on the TM stability and the micromechanics involved in the sound -evoked motion of stereocilia might be involved in familial MD.

Genomics, Hearing loss, Inner ear, Meniere’s disease, Vestibular disorders . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint

The tectorial membrane (TM) is an extracellular matrix with a unique structure1,2 containing three different collagens and three non-collagenous glycoproteins expressed at high levels just in the inner ear: α-tectorin, β-tectorin, and otogelin3–6. Several roles have been suggested for the TM in hearing such as contributing to control of hearing sensitivity by influencing the ionic environment around the hair cell stereocilia7. Sound stimulates movement of the hair cells relative to the TM, deflects the stereocilia, and leads to oscillations in the hair cell membrane potential, transducing sound into electrical signals. Alpha-tectorin (encoded by TECTA) is one of the main non -collagenous proteins of the TM , an extracellular matrix that lies over the stereocilia of the hair cells and it is critical for the deflection of the stereocilia and gating of mechanotransduction channel s8. Mutations in this gene, located in 11q22 -q24, are known to cause two phenotypes of non -syndromic autosomal hereditary hearing loss 9: dominant (Deafness, Autosomal Dominant 8/12; DFNA8/12 )10 and recessive (Deafness autosomal recessive 21; DFNB21)11. To date, over 122 pathogenic or likely pathogenic variants have been identified in the TECTA gene that may cause non -syndromic hearing loss (NSHL), not being associated with any other symptoms (http://deafnessvariationdatabase.org/; Release: 4 Jan 2021; Accessed: 29 Oct 2021)12. DFNA8/12, first described in 1998 by Kirs chhofer et al.8, showed an Austrian family in which 11 individuals covering 4 gen erations had autosomal dominant NSHL. It was defined by sensorineural hearing loss (SNHL) affecting a broad spectrum of frequencies depending on the domain in which the mutation occurs 13 with a typical U -shaped audiogram. Its onset could be prelingual or in childhood and the progression of hearing loss can be stable or progressive. TECTA mutations are reported in up to 4% of all cases of autosomal dominant NSHL14. Mustapha et al. (1999), described a Lebanese family belonging to the Shiite community in which 9 members presented with a prelingual severe-to-profound sensorineural isolated form of . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint deafness11. DFNB21 is defined by a severe-to-profound prelingual hearing loss (70-110dB on all frequencies, most pronounced in mid -frequencies) with a cookie -bite or U -shape audiometric configuration15. Nonsense mutations that cause DFNB21 create premature stop codon s that truncate the α-tectorin protein, leading to a non-functional version of the protein. A total loss of α- tectorin function alters the structure of the TM in such a way that sound cannot be converted to nerve impulses16. Meniere disease (MD [OMIM 156000]) is an inner ear disorder with an estimated prevalence of 75/100,000 in the Southern European population17. It is defined by episodes of vertigo associated with SNHL, initially affecting low- to medium frequencies, tinnitus , and aural fullness 18. Familial aggregation in MD has been described in 9 -10% of European descendant population 19. Most families show an autosomal dominant inheritance pattern an d rare missense variants in several genes have been described in unrelated MD families 20,21. Nevertheless, none of these variants has been reported in other MD families, supporting genetic heterogeneity in familial MD. Recently in two different studies, enrichment of rare missense variants in the OTOG gene was found in 15 unrelated MD families22 and, in an exome sequencing study in 62 MD families, other 9 families showed rare heterozygous variants in the MYO7A gene associated with the MD phenotype23. Here, we have performed bioinformatic analyses in exome sequencing data obtaine d from 99 individuals (77 families with MD) and found 7 patients in 4 unrelated families with rare missense variants and a short deletion in the TECTA gene. Variants in this gene were also found in two additional families with one MD patient and relatives with partial syndromes carrying a missense heterozygous variant and a short deletion. We suggest that these deletions and missense variants in the TECTA gene could change the TM micromechanics involved in the sound -evoked motion of stereocilia in familial MD. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint

& METHODS Ethics This study protocol was approved by the local eth ics committees (MS/2014/02, Institutional Review Board for Clinical Research, Universidad de Granada, Spain; KEK -ZH-Nr. 2019-01006, Kantonale Ethikkommission, Zurich, Switzerland) and a written informed consent to donate biological samples was obtained fro m all subjects. The study was carried ou t according to the principles of the Declaration of Helsinki revised in 2013 for investigation with human subjects. Patient assessment, selection, and clinical phenotyping MD patients were diagnosed and recruited from different Spanish hospitals within th e Meniere’s Disease Consortium (MeDiC). Patients were diagnosed according to the diagnostic criteria defined by the International Classification Committee for Vestibular Disorders of the Barany Society in 201518. A complete audiological and vestibular assessment was performed, including magnetic resonance imaging in all cases, to rule out other vestibular diseases that could explain the phenotype. Pure-tone audiograms were retrieved to assess hearing loss since the onset of the disease. A total of 99 Spanish patients with familial MD over 18 years old from 77 different families were selected for exome sequencing. DNA extraction and exome sequencing Blood or saliva samples were collected to obtain DNA as previously described 24. DNA samples were extracted with QIAamp DNA Mini Kit (Qiagen, Venlo, The Netherlands) and prepIT -L2P (DNA Genotek, Ottawa, Canada), respectively, following the manufacturer’s protocols. DNA concentration and quality para meters were checked by Nanodrop (Thermofisher) and Qubit (Invitrogen) to assess that the samples reach the quality and concentration required for exome sequencing. Additionally, DNA integrity was verified by electrophoresis in a 2% agarose gel. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Exome sequencing was chosen for this study. All exomes were sequenced as previously described23. Exome libraries and coding regions were selected using the Agilent SureSelect XT v6 Exome kit (Agilent Technologies, Santa Clara, CA, USA). Sequencing was performed in a NovaSeq 6000 platform (Illumina) with a mean coverage of 100X. Processing and dataset generation All sequenced samples were aligned using the GRCh38/Hg38 reference genome with the maximal exact matches algorithm from Burrows-Wheeler Aligner. For variant calling, we followed standard recommended criteria from GATK. Exome reference a lignment, base quality score recalibration, variant calling, and quality filtering pipeline was addressed using Sarek Nextflow pipeline (NF-ACore)25. Post-alignment processing was used to remove duplicated reads and the quality of the alignment itself was assessed 26, genetic variants were then called using the Haplotypecaller function from GATK. After the calling, we merged all the files to generate the MD variant dataset. We performed a variant quality filtering step using the va riant quality score recalibration approach recommended by GATK. As a result, two variant call format files were generated, retrieving single nucleotide variants (SNVs) and short insertion and deletions from the sequenced exomes. Annotation and prioritization strategy To annotate the MD variant dataset, we used the Variant Effect Predictor (VEP, Ensembl). Then, we selected minor allele frequency ( MAF) thresholds of 0.005 and 0.0005 to identify both autosomal-recessive and autosomal -dominant rare variants in the familial MD cohort, respectively, based on the data from a multi -ethnic study that assessed the pathogenicity of reported NSHL variants27. Allelic frequencies were retrieved for the non-Finish European (NFE) population from the Exome Aggregation Consortium (ExAC; N=32,299) and the Genome . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Aggregation Database (gnomAD; N=33,365) databases, and the Spanish population from the Collaborative Spanish Variant Server (CSVS; N=1,942) database28. Candidate variants were then classified according to the American College of Medical Genetics (ACMG) an d the Association for Molecular Pathology (AMP) guidelines 29, according to the specific guidelines for variant interpretation in genetic hearing loss30. Additionally, in-silico tools such as Sorting Intolerant From Tolerant (SIFT; SIFT0.446), Genomic Evolutionary Rate Profiling (GERP; GERP>2) or Combined Annotated Dependent Depletion (CADD; CADD>15) were used to prioritize and classify each variant according to its predicted pathogenicity. Lastly, as an additional variant prioritization method, candidate genes carrying rare variants were associated with mammalian phenotypes using the Mouse Genome Database (http://www.informatics.jax.org). Similarly, the Human Phenotype Ontology Project (https://hpo.jax.org/app/) and the Online Mendelian Inheritance in Man (OMIM; https://omim.org/) databases were used to determine associations in humans between candidate genes and phenotypes. Statistical analysis European and Spanish databases such as ExAC, gnomAD, and CSVS were used as references to compare the observed MAF in familial MD. Odds ratios with 95% confidence interval were calculated for each single or set of variants. One-side p-values were corrected for multiple testing following the Bonferroni approach. A corrected p-value < 0.05 was considered statistically significant. The same approach was followed to assess if a combination of variants showed a significant association. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Candidate variant validation and representation Regions with prioritized variants were visuall y inspected using the Integrative Genome Viewer software. Novel variants were validated using Sanger sequencing. Primers used for PCR (Supplementary Table 1) were designed neighboring the regions flanking the variants using the Primer3 v4.1 (http://bioinfo.ut.ee/primer3/), Primer-Blast (https://www.ncbi.nlm.nih.gov/tools/primer-blast/), and the Oligoanalyzer tool (https://eu.idtdna.com/calc/analyzer/). Candidate genes and variants were represented using Illustrator for Biological Sequences Version 1.031. Hearing assessment and analysis To analyze the time course of the hearing profile in familial MD cases with candidate variants, standard audiometric evaluations for air and bone conduction prompted by pure tones from 125 to 8,000Hz were retrieved from their clinical records. Αlpha-tectorin protein model The effect of candidate variants on the α -tectorin protein structure was evaluated by protein modeling. The human α -tectorin mature amino acid sequence (from 23 to 2 ,091) was retrieved from Uniprot entry O75443, containing 10 functional domains. Structural models of the domains were generated using MODELLER (homology modeling)32, Robetta-ab33 (ab-initio modeling), and AlphaFold234 (ab-initio modeling) methods, and the best ones were assembled using the DEMO method35. Each of the generated models and the final assembly were validated using the structure validation algorithms MolProbity36, Verify3D37, ERRAT38, ProSA-Web39, and QMEANDisCo40. The in-silico model was used to predict the stability change (ΔΔG) of the α-tectorin produced by the candidate variants. Hence, we used DynaMut2 41, MAESTROweb 42, mCSM 43, PremPS 44 and CUPSAT45 tools. Variants were classified as neutral when -0.5 < ΔΔG < 0.546. Finally, a prediction of Ca2+ ion binding sites was also performed to observe the possible effects of the variants on Ca2+ uptake with the metal ion-binding site prediction (MIB) method47. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Variant data and protein structural model submission All candidate variants in the TECTA gene have been submitted to the Clinvar database (http://www.ncbi.nlm.nih.gov/clinvar/). The accession numbers for these variants are as follow: SUB10993075, SUB10994469 and SUB10994413. The α-tectorin structural model was submitted to the ModelArchive database (https://modelarchive.org/doi/10.5452/ma-xd6ic; Public access after publication. Temporary access code: mwnV7cphoi) . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint

Rare missense variants and deletions in the TECTA gene in MD A total of 34,833 and 20,684 nonsynonymous (missense, start-loss, stop-gain, and stop-loss) or splice-site variants were considered after using, respectively, MAF thresholds of 0.005 and 0.0005 in the familial MD cohort. These rare variants were found in 9,666 genes using the most restrictive MAF threshold and in 12,366 genes using the 0.005 MAF threshold. After assessing and analy zing those genes with non -singleton variants (i.e., observed in more than one patient), and beyond the genes related to familial MD already published22,23, the TECTA gene was highlighted as the candidate for 4 multicase MD families carrying rare missense heterozygous variants (F1 – F3) and a short deletion (F4) in this gene. Variants in this gene were also found in two additional families with one MD patient and relatives with partial syndromes carrying a missense heterozygous variant (F5) and a short deletion (F6) (Table 1). The three variants in all multicase MD families were found clustered in the zonadhesin-like region, between the third trypsin inhibitor -like cysteine-rich (TIL) domain and the fourth von Willebrand factor type D (VWFD) domain. On the other hand, the two variants observed in the two families with partial syndromes were found close to the zona pellucida (Figure 1). Of note, the variant chr11:121158016T>C (p .Val1494Ala) was found in F1, F2 , and a sporadic case. The segregation of this variant was confirmed in F1, where it segregated in all three affected MD individuals and the father of III-7, who only suffered from hearing loss. The frequency of this variant in the NFE population from gnomAD is 8.8x10-5 and no individuals have been reported as homozygous for the alternate allele. This variant is predicted to impact the protein by several in- silico tools (GERP=3.09, PolyPhen=0.954 , CADD=24.3) and it was classified as a variant of unknown significance (VUS). Interestingly, a rare variant in chr10:112298172A>C (p.Lys259Gln) in the TECTB was found in F2 along with the p .Val1494Ala TECTA variant. Moreover, a novel . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint frameshift deletion in chr11:121157956AC>A (p.Asn1474LysfsTer91) found in F4 was classified as likely pathogenic. This deletion resulted in a premature stop codon after 91 amino acids that generates a truncated protein of 1594 residues instead of 2155. The missense variant chr11:121152980G>C (p.Cys1402Ser) found in F3, whose frequency in the NFE pop ulation is 1.5x10-5, was also predicted to impact the protein (SIFT=0, GERP=5.33, PolyPhen=0.993; CADD=28), and it was classified as VUS according to the ACMG/AMP guidelines. The two additional variants observed in the families with partial syndromes were classified as VUS. The variant chr11:121165368C>T (p .Pro1790Ser) was found in F5 segregating in two patients, the father with MD and h is daughter with only vestibular symptoms. This is a novel variant not observed in gnomAD. Finally, the short deletion chr 11:121189864GC>C (p.Gly2118ProfsTer22) observed in F6 results in a premature stop codon at position 2139, generating a slightly shorter protein . This variant has been observed in only one individual (African/African American population) from gnomAD. Clinical description of families carrying variants in the TECTA gene A briefcase report was made of each family to assess the genotype -phenotype correlations. Figures 2 and 3 show the pedigrees and the pure -tone audiograms, respectively, of the six families included in this study. Supplementary Table 2 shows a summary of the clinical information of familial MD patients carrying variants in the TECTA gene. Family 1 (F1) A family consisting of three women in two generations with the complete phenotype was selected for exome sequencing. In this family, six relatives had incomplete phenotypes: five relatives (four women and one man) with recurrent vertigo and one man with SNHL. Additionally, three relatives (two women and one man) also suffered from migraine. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint The index case (III-11) was a woman suffering from right SNHL with episodes of vertigo since her 40s. However, in few years, the hearing loss evolved to bilateral SNHL. Her parent (II-6), who is currently in her 80s, was diagnosed with left-sided MD when she was 46-50 years old. The last time she was examined the patient presented bilateral hearing loss with a normal gain in the video head impulse test (vHIT). Finally, the relative III-7 was diagnosed with bilateral MD. Currently, III- 7 is 5 1-55 years old and, when she was evaluated , she did not show hearing nor vestibular symptoms, showing normal hearing and vestibulo-ocular reflexes. Family 2 (F2) This family consists of two female cousins with MD in one generation (III-2 and III-6) where only III-6 (index patient) was available for exome sequencing. III-6 is in her 70s a nd suffers from unilateral (left ear) hearing loss since she was 5 1-55. The disease progressed from unilateral to bilateral MD when she was 6 1-65 years old, suffering from that moment mainly vestibular symptoms. Some of her relatives had partial syndromes: her parent (II-6) suffered from tinnitus, a trait that is also present in her child (IV-4). In addition, one of her relatives (II-8) presented SNHL and vertigo episodes who did not meet the diagnostic criteria for MD. Family 3 (F3) The third family consists of two relatives with MD, the parent (I-2) and her child (II-3), only being available for exome sequencing II-3. There are no intermediate phenotypes in this family, neither hearing loss, nor vertigo or tinnitus. Two relatives of II-3 have a history of high blood pressure. Additionally, one of them was diagnosed with cardiac valvulopathy and the other with psoriasis. The index case (II-3) is a woman currently in her 50s suffering from fluctuating hearing loss and tinnitus in her right ear since she was 41-46. She experienced her first vertigo episode associated with fluctuating hearing loss at 45 yea rs old. During the next 5 months, she suffered from 8 episodes of vertigo, 4 of them lasting more than two hours. Right canal paresis (62%) was . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint evidenced by bithermal caloric testing. The patient started treatment with acetazolamide (250 - 500 mg daily) without any improvement in her symptoms. Next, two intratympanic instillations of dexamethasone (23mg/ml) were administrated in the right ear, with an improvement in control of vertigo. In a 10 -year-follow-up, this patient suffered from 4 episodes of vertigo. Her parent (I-2) suffered from left-sided MD, but it progressed from unilateral to bilateral MD with severe hearing loss at the end of the disease. This patient used a combination of trimetazidine and vestibular sedatives to control vertigo attacks. Family 4 (F4) A fourth family consisting of two relatives with MD in two generations ( parent and child) was studied. Only the child (II-1, index patient) was available for exome sequencing. Her parent, who died at the age of 36-40, suffered from unilateral MD since her thirties. II-1 is currently a woman in her 80s suffering from bilateral MD. At the age of 21-25, the disease started as unilateral, affecting her right ear. We could not assess the hearing loss progression in this ear since the first audiogram we could recover was when the patient already suffered from severe SNHL affecting all frequencies. Over the years, the disease progressed affecting both ears, estimating the onset of the left hearing loss between the ages of 40 and 50. Hearing los s progressed in both ears showing a flat -type audiogram in her 70s . The vestibular assessment was performed at 76-80 year old by vHIT testing, showing a bilateral vestibular hypofunction. Due to a tympanostomy tube, caloric testing could not be performed. In this same year, the patient got a cochlear implant in her right ear. Family 5 (F5) A family consisting of a man with MD ( II-3; index patient ) and his child (III-1) with partial syndromes. II-3 is a man in his 70s suffering from bilateral MD since he was 31-35. We could not evaluate the hearing progression on this patient since the first audiogram we were able to retrieve . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint was when he was 6 1-65 and the disease had progressed for more than 30 years. At that time, hearing loss was already affecting all frequencies in both ears. His child, currently in her 50s , reported episodes of vertigo accompanied by tinnitus and headache during the last 5 years. However, she did not show a hearing loss. None of the other relatives reported MD nor partial syndromes: II -1 and III -2 suffered from high blood pressure and II -2 suffered from diabetes mellitus type 2. Family 6 (F6) The last family consisted of a woman with MD (IV-3) and two other relatives with partial syndromes (a woman (I-1) and a man (II-7)). The index patient (IV -3) is a woman in her 50s suf fering from recurrent vertigo associated with tinnitus and aural fullness (left ear) since she was 2 6-30. This patient suffers from profound unilateral hearing loss affecting all frequencies since the first years of the disease. According to the familial pedigree, these traits (i.e., hearing loss or vertigo) were exclusively identified on the maternal side of the proband: her relative I-2 suffered from unknown etiology hearing loss and vertigo and her relative II-7 only suffered from hearing loss. Sporadic patient We have also identified a patient (woman; 71-75 years old) with sporadic MD carrying the same missense variant in the TECTA gene as F1 and F2 patients. This patient reported her first episode of vertigo when she was 61-65, when she had two prolonged episodes of vertigo, referring to a sensation of spinning motion, right-sided hearing loss, and vegetative symptoms. One week after her second vertigo episode, the audiogram showed normal thresholds except for a slight drop in high frequencies. Few years later, the patient returned to the hospital reporting a new episode of vertigo, beginning with a sensation of spinning motion followed by tinnitus and hearin g loss in her right ear. At that time, the audiogram showed a moderate to severe SNHL in her right ear involving all frequencies. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint During that year, the patient reported new episodes of vertigo. Caloric testing was conducted, showing normal results. Vertigo attacks were controlled with steroids and betahistine. When she was 71-75, the patient reported multiple episodes of vertigo lasting several hours with worsening hearing loss and aural fullness. The patient followed her treatment of betahistine and a salt-reduced diet was recommended. During th e last years, the patient has not reported any vestibular symptoms, although she has referred tinnitus in her right ear. Protein modeling The α-tectorin protein model was obtained by modeling the protein domains using the AlphaFold2

and assembled with the DEMO method (Figure 4). According to the geometrical validation

( Supplementary Table 3) , we have obtained a reliable model compared to structures solved by experimental techniques at the geometric level. This model was used to predict the impact of SNVs on protein stability and Ca2+ binding sites. p.Cys1402Ser (Cys1380Ser, in the mature protein without the signal peptide and the final propeptide), p.Val1494Ala (Val1472Ala) , and p.Pro1790Ser (Pro1768Ser) variants were predicted in-silico to change the overall stability of the α -tectorin (Supplementary Table 5). Consistent with the used methods, variants p.Cys1402Ser found in F3 and p.Val1494Ala found in F1 and F2 were classified as destabilizing variants. In contrast, most of the used p redictors classified the novel variant p.Pro1790Ser found in F5 as a neutral variant according to the predicted perturbation on protein stability. Four probable Ca2+ binding sites were predicted in the α-tectorin VWFD domain where the p.Pro1790Ser is located, however, this variant appears to not affect them (Supplementary Table 4 and Supplementary Figure 2). . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint

Our findings support those rare variations in the TECTA gene may affect the stability of the protein α-tectorin in patients with familial MD suggesting a novel mechanism involving the TM in MD. The TM is an acellular structure that covers the sensory epithelium of the cochlea in the organ of Corti. This membrane, which is composed of a mixture of collagenous and non -collagenous proteins, plays a critical role in the process of hearing mediating the mechanical stimulation of cochlear hair cells48. Thus, variants in these non-collagenous proteins of the TM, namely otogelin (OTOG), otogelin-like ( OTOGL), α -tectorin ( TECTA), β -tectorin ( TECTB), and the carcinoembryonic antigen cell adhesion molecule 16 ( CEACAM16), may result in different types of autosomal dominant or autosomal recessive hearing loss in humans: DFNB18B49, DFNB84B50, DFNA8/1210, DNFB2111, DFNA4B51 or DFNB11352. In this study, we have identified 3 rare missense variants and 2 deletions in the TECTA gene in 4 multicase MD families and 2 families with MD and partial syndromes by exome sequencing. Of note, one of these rare variants (p.Val1494Ala) was found in two unrelated families segregating the MD phenotype, and two deletions generating a truncated form of the α-tectorin protein in other two families. These variants in α-tectorin could change the TM stability, disturbing the sound-evoked motion of hair cell stereocilia in familial MD. Familial aggregation is reported in 9-10% of MD patients showing, mostly, an autosomal dominant inheritance pattern19. However, other inheritance patterns have been recently proposed involving rare variants in OTOG and MYO7A genes. Enrichment of rare variants was found in the OTOG gene in 15 families with MD, suggesting a recessive inheritance pat tern mediated by compound heterozygous variants in this gene22. The OTOG gene encodes otogelin, an extracellular protein that participates in the otolith tethering in the otolithic membrane, the TM attachment crowns, and the horizontal top connectors between stereocilia 6,53. On the other hand, it was suggested a potential digenic inheritance pattern in familial MD mediated by rare missense variants in MYO7A and other genes encoding for proteins involved in the organization of the stereocilia links (e.g., . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint CDH23, PCDH15, or ADGRV1)23. The MYO7A gene encodes the myosin VIIA protein, a motor protein involved in the formation of ankle links and tips links in the stereocilia of hair cells54. These two genes, OTOG and MYO7A, are closely related to the TECTA gene since they are highly expressed in the hair cell stereocilia and involved in the interaction with the TM. Rare variants in any of these genes could reduce the stability of the stereocilia proteins, the attachments of the apical crowns of the stereocilia to the TM or generate structural changes in the TM. This may lead to a sudden onset of spontaneous oscillations in a subset of hair bundles, resulting in a spurious signal that may cause abnormal depolarization of hair cells and fluctuation in hearing in MD. The TECTA gene encodes the α -tectorin protein, one of the major non -collagenous proteins of the TM48. This large p rotein contains 10 functional domains divided into three major regions: 1) an entactin-like (NIDO) region; 2) a larger middle region, the zonadhesin region containing a von Willebrand factor type C (VWFC) domain, four VWFD domains , and three TIL domains; a nd 3) the zona pellucida region5. Although there is little evidence, several molecular models involving α-tectorin, β-tectorin, and CEACAM16 have been proposed for the formation of the striated-sheet matrix, strands of filaments that organize the collagen fibers in the TM55,56. Both, α- and β-tectorin contain a ZP region, a polymerization domain that could mediate the formation of either homomeric or heteromeric filaments crosslinked by CEACAM16. Studies in mice lacking functional TECTA (TectaΔENT/ΔENT) showed a TM absent of striated -sheet matrix and completely detached from the organ of Corti16. Furthermore, several models carrying heterozygous variants in different Tecta domains were generated as models from human TECTA variants. These variants disrupted the structure of the covernet fibrils, the marginal band, the Hensen’s stripe, and the Kimura’s membrane57,58. The structural consequences of variants in the TECTA gene showed in this study are expected to be milder compared to the phenotype showed in the TectaΔENT/ΔENT mice model, resembling more t he phenotypes exhibited by those mice models carrying heterozygous variants in Tecta. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint The SNVs described in this study were investigated at the functional level according to the predicted α-tectorin protein model. The variant p.Val1494Ala was found in F1, F2 and a sporadic case of our sporadic MD cohort, being classified as VUS according to the ACMG /AMP criteria. Although this variant was located in the VWFD domain, a domain that can bind Ca2+, the variant p.Val1494Ala seems to have no effect in the Ca2+ binding sites predicted using the protein model structure. In addition, based on the protein stability predictors, this variant produces a slightly destabilizing effect on α-tectorin. The missense variant p.Cys1402Ser found in F3 was classified as a VUS according to the ACMG /AMP guidelines. The change of cysteine to serine at residue 1402 can lead to the breaking of the Cys1359-Cys1402 disulfide bond. The suggested disulfide bond disruption could destabilize the striated -sheet matrix structure of the TM, causing progressive hearing loss59. In addition, a novel frameshift deletion p .Asn1474LysfsTer91 was found in F4 , resulting in the absence of the fourth VWFD domain and the ZP domain. ZP and VWFD domains have been suggested to be involved in the TM matrix assembly through the formation of homomeric filaments of α-tectorin or heteromeric filaments with β-tectorin5,60. In addition, the TM has been suggested as a reservoir for Ca2+ cations, which are needed to control the mechanotransduction channel in the stereocilia of hair cells7. Apparently, behind this function are the VFWD domains of α-tectorin and otogelin proteins, which can bind Ca 2+ ions61. Thus, the short deletion found in F4 could potentially decrease Ca 2+ uptake by the VWFD domain and prevent the formation of α -tectorin homodimers and α-tectorin/β-tectorin heterodimers. Rare variants in the TECTA gene were also detected in two families with MD and partial syndromes. The novel variant p.Pro1790Ser in F5 was classified as a neutral variant in terms of protein stability perturbation by 4 out of 6 prediction tools. Neverthele ss, we cannot discard a deleterious effect of this variant for a reason different from stability. In fact, this variant, which affects the ZA -ZP interdomain of α -tectorin, is only one amino acid residue downstream of the . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint p.Pro1791Arg variant, firstly described by Hildebrand et al. in an American family suffering from prelingual mid -frequency SNHL 14. Later, this variant was also found in a man suffering from progressive, postlingual SNHL62. Finally, although the p .Gly2118ProfsTer22 found in F6 occurs in the last exon, which is part of the α-tectorin propeptide, this region contains elements essential for its function. The α-tectorin suffers a post-translational modification in which it is tethered to the membrane via glycosylphosphatidylinositol (GPI), a process that seems to be needed to prevent the diffusion of secreted TM component. The C-terminal, region where the short deletion found in F6 is located, contains the GPI anchorage signal 63. As a result, this frameshift deletion could be involved in an alteration of the TM during its formation by the modification of the GPI anchorage signal and leading to the clinical phenotype. The phenotype observed in the six fam ilies carrying rare heterozygous variants in the TECTA gene was described. In general, we observed a trend towards postlingual hearing loss worsening with age involving all frequencies. Conversely, we did not identify a different audiometric profile according to the region where the variants were localized, nor the cookie -bite/U-shaped audiogram typically observed in patients carrying recessive mutations in TECTA9,15,64. Interestingly, the patients carrying frameshift deletions in this study (i.e., F4 & F6) presented an earlier age of onset compared to those patients carrying rare missense variants, suggesting a more significant alteration in the TM matrix that could cor relate to an earlier occurrence of the phenotype. Unlike the hearing loss phenotype, the association between variants in the TECTA gene and vestibular dysfunction has not been established since most patients carrying variants in α-tectorin do not show a vestibular phenotype. Nevertheless, some of these patients reported episodic vertigo or showed vestibular hyporeflexia 60,65,66. Furthermore, the expression of α-tectorin in the vestibular system has been demonstrated in several mouse models, being detected in the saccule and the utricule67. Mice lacking α-tectorin showed reduced otoconial membranes with no obvious . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint behavioral defects 16. In addition, a mechanism for otolith tethering requiring otogelin and α - tectorin has bee n proposed in a zebrafish model 53. In humans, Calzada et al . found that all patients who experienced drop attacks showed disrupted utricular otolithic membranes. Additionally, they showed that the average thickness of the utricular otolithic membrane in patients with MD was 11.45 micrometers, while in normal tissue the mean thickness was 38 micrometers68. This reduced thickness in the otolithic membrane could be associated with changes in the structure of the proteins involved in the otolithic membrane, including α -tectorin. Nevertheless, since the role of α -tectorin in the human vestibular system is not completely clear, we cannot discard other genes or epigenetic factors that could modulate the vestibular phenotype in MD patients. The results achieved in this study seem to be in line with the results obtained in most recent genetic studies about familial MD, where rare variants in MYO7A and OTOG gene were suggested to modify the stability or the interactions of different proteins in the apical surface of the sensory epithelia, such as hair cells (stereocilia) or the TM22,23. In this study, the presence of rare missense variants and frameshift deletions in the TECTA gene in six unrelated families with MD suggests a role of this gene in the pathophysiology of the disease. However, because of the lack of a reliable association between α -tectorin and the vestibular function, we consider that additional variants and genes may contribute to the vestibular phenotype in MD. DATA AVAILABILITY The data that support the findings of this study are available from the corresponding author upon reasonable request. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint LIST OF ABBREVIATIONS ΔΔG: Stability change ACMG: American College of Medical Genetics AMP: Association for Molecular Pathology CADD: Combined Annotated Dependent Depletion CSVS: Collaborative Spanish Variant Server ExAC: Exome Aggregation Consortium GERP: Genomic Evolutionary Rate Profiling gnomAD: Genome Aggregation Database GPI: Glycosylphosphatidylinositol MAF: Minor allele frequency MD: Meniere’s disease MeDiC: Meniere’s Disease Consortium MIB: Metal ion-binding site prediction NFE: Non-Finish European NSHL: Non-syndromic hearing loss Polyphen: Polymorphism Phenotyping SIFT: Sorting Intolerant From Tolerant SNHL: Sensorineural hearing loss SNVs: Single nucleotide variants TIL: Trypsin Inhibitor like cysteine-rich TM: Tectorial membrane VEP: Variant Effect Predictor vHIT: video head impulse test VWFC: von Willebrand factor type C VWFD: von Willebrand factor type D ZP: Zona-pellucida . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint

1. Lim DJ. Functional structure of the organ of Corti: a review. Hear Res. 1986;22:117-146. doi:10.1016/0378-5955(86)90089-4 2. Hasko JA, Richardson GP. The ultrastructural organization and properties of the mouse tectorial membrane matrix. Hear Res. 1988;35(1):21-38. doi:10.1016/0378- 5955(88)90037-8 3. Richardson GP, Russell IJ, Duance VC, Bailey AJ. Polypeptide composition of the mammalian tectorial membrane. Hear Res. 1987;25(1):45-60. doi:10.1016/0378- 5955(87)90078-5 4. Thalmann I, Thallinger G, Crouch EC, Comegys TH, Barrett N, Thalmann R. Composition and supramolecular organization of the tectorial membrane. The Laryngoscope. 1987;97(3 Pt 1):357-367. 5. Legan PK, Rau A, Keen JN, Richardson GP. The mouse tectorins. Modular matrix proteins of the inner ear homologous to components of the sperm-egg adhesion system. J Biol Chem. 1997;272(13):8791-8801. doi:10.1074/jbc.272.13.8791 6. Cohen-Salmon M, El-Amraoui A, Leibovici M, Petit C. Otogelin: A glycoprotein specific to the acellular membranes of the inner ear. Proc Natl Acad Sci U S A. 1997;94(26):14450- 14455. 7. Strimbu CE, Prasad S, Hakizimana P, Fridberger A. Control of hearing sensitivity by tectorial membrane calcium. Proc Natl Acad Sci U S A. 2019;116(12):5756-5764. doi:10.1073/pnas.1805223116 8. Kirschhofer K, Kenyon JB, Hoover DM, et al. Autosomal-dominant, prelingual, nonprogressive sensorineural hearing loss: localization of the gene (DFNA8) to chromosome 11q by linkage in an Austrian family. Cytogenet Cell Genet. 1998;82(1- 2):126-130. doi:10.1159/000015086 9. Balciuniene J, Dahl N, Jalonen P, et al. Alpha-tectorin involvement in hearing disabilities: one gene--two phenotypes. Hum Genet. 1999;105(3):211-216. doi:10.1007/s004390051091 10. Verhoeven K, Van Laer L, Kirschhofer K, et al. Mutations in the human alpha-tectorin gene cause autosomal dominant non-syndromic hearing impairment. Nat Genet. 1998;19(1):60- 62. doi:10.1038/ng0598-60 11. Mustapha M, Weil D, Chardenoux S, et al. An alpha-tectorin gene defect causes a newly identified autosomal recessive form of sensorineural pre-lingual non-syndromic deafness, DFNB21. Hum Mol Genet. 1999;8(3):409-412. doi:10.1093/hmg/8.3.409 12. Azaiez H, Booth KT, Ephraim SS, et al. Genomic Landscape and Mutational Signatures of Deafness-Associated Genes. Am J Hum Genet. 2018;103(4):484-497. doi:10.1016/j.ajhg.2018.08.006 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint 13. Sagong B, Park R, Kim YH, et al. Two novel missense mutations in the TECTA gene in Korean families with autosomal dominant nonsyndromic hearing loss. Ann Clin Lab Sci. 2010;40(4):380-385. 14. Hildebrand MS, Morín M, Meyer NC, et al. DFNA8/12 caused by TECTA mutations is the most identified subtype of nonsyndromic autosomal dominant hearing loss. Hum Mutat. 2011;32(7):825-834. doi:10.1002/humu.21512 15. Naz S, Alasti F, Mowjoodi A, et al. Distinctive audiometric profile associated with DFNB21 alleles of TECTA. J Med Genet. 2003;40(5):360-363. doi:10.1136/jmg.40.5.360 16. Legan PK, Lukashkina VA, Goodyear RJ, Kössi M, Russell IJ, Richardson GP. A targeted deletion in alpha-tectorin reveals that the tectorial membrane is required for the gain and timing of cochlear feedback. Neuron. 2000;28(1):273-285. doi:10.1016/s0896- 6273(00)00102-1 17. Morales Angulo C, Gómez Castellanos R, García Mantilla J, Bezos Capelastegui JT, Carrera F. Epidemiología de la enfermedad de meniere en cantabria. Acta Otorrinolaringológica Esp. 2003;54(9):601-605. doi:10.1016/S0001-6519(03)78456-8 18. Lopez-Escamez JA, Carey J, Chung WH, et al. Diagnostic criteria for Menière’s disease. J Vestib Res Equilib Orientat. 2015;25(1):1-7. doi:10.3233/VES-150549 19. Requena T, Espinosa-Sanchez JM, Cabrera S, et al. Familial clustering and genetic heterogeneity in Meniere’s disease. Clin Genet. 2014;85(3):245-252. doi:10.1111/cge.12150 20. Escalera-Balsera A, Roman-Naranjo P, Lopez-Escamez JA. Systematic Review of Sequencing Studies and Gene Expression Profiling in Familial Meniere Disease. Genes. 2020;11(12):E1414. doi:10.3390/genes11121414 21. Gallego-Martinez A, Lopez-Escamez JA. Genetic architecture of Meniere’s disease. Hear Res. 2020;397:107872. doi:10.1016/j.heares.2019.107872 22. Roman-Naranjo P, Gallego-Martinez A, Soto-Varela A, et al. Burden of Rare Variants in the OTOG Gene in Familial Meniere’s Disease. Ear Hear. 2020;41(6):1598-1605. doi:10.1097/AUD.0000000000000878 23. Roman-Naranjo P, Moleon MDC, Aran I, et al. Rare coding variants involving MYO7A and other genes encoding stereocilia link proteins in familial meniere disease. Hear Res. 2021;409:108329. doi:10.1016/j.heares.2021.108329 24. Szczepek AJ, Frejo L, Vona B, et al. Recommendations on Collecting and Storing Samples for Genetic Studies in Hearing and Tinnitus Research. Ear Hear. 2019;40(2):219-226. doi:10.1097/AUD.0000000000000614 25. Garcia M, Juhos S, Larsson M, et al. Sarek: A portable workflow for whole-genome sequencing analysis of germline and somatic variants. F1000Research. 2020;9:63. doi:10.12688/f1000research.16665.2 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint 26. McKenna A, Hanna M, Banks E, et al. The Genome Analysis Toolkit: a MapReduce framework for analyzing next-generation DNA sequencing data. Genome Res. 2010;20(9):1297-1303. doi:10.1101/gr.107524.110 27. Shearer AE, Eppsteiner RW, Booth KT, et al. Utilizing ethnic-specific differences in minor allele frequency to recategorize reported pathogenic deafness variants. Am J Hum Genet. 2014;95(4):445-453. doi:10.1016/j.ajhg.2014.09.001 28. Peña-Chilet M, Roldán G, Perez-Florido J, et al. CSVS, a crowdsourcing database of the Spanish population genetic variability. Nucleic Acids Res. 2021;49(D1):D1130-D1137. doi:10.1093/nar/gkaa794 29. Richards S, Aziz N, Bale S, et al. Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med Off J Am Coll Med Genet. 2015;17(5):405-424. doi:10.1038/gim.2015.30 30. Oza AM, DiStefano MT, Hemphill SE, et al. Expert specification of the ACMG/AMP variant interpretation guidelines for genetic hearing loss. Hum Mutat. 2018;39(11):1593-1613. doi:10.1002/humu.23630 31. Liu W, Xie Y, Ma J, et al. IBS: an illustrator for the presentation and visualization of biological sequences. Bioinforma Oxf Engl. 2015;31(20):3359-3361. doi:10.1093/bioinformatics/btv362 32. Webb B, Sali A. Protein Structure Modeling with MODELLER. Methods Mol Biol Clifton NJ. 2021;2199:239-255. doi:10.1007/978-1-0716-0892-0_14 33. Kim DE, Chivian D, Baker D. Protein structure prediction and analysis using the Robetta server. Nucleic Acids Res. 2004;32(Web Server issue):W526-531. doi:10.1093/nar/gkh468 34. Jumper J, Evans R, Pritzel A, et al. Highly accurate protein structure prediction with AlphaFold. Nature. 2021;596(7873):583-589. doi:10.1038/s41586-021-03819-2 35. Zhou X, Hu J, Zhang C, Zhang G, Zhang Y. Assembling multidomain protein structures through analogous global structural alignments. Proc Natl Acad Sci U S A. 2019;116(32):15930-15938. doi:10.1073/pnas.1905068116 36. Williams CJ, Headd JJ, Moriarty NW, et al. MolProbity: More and better reference data for improved all-atom structure validation. Protein Sci Publ Protein Soc. 2018;27(1):293-315. doi:10.1002/pro.3330 37. Eisenberg D, Lüthy R, Bowie JU. VERIFY3D: assessment of protein models with three- dimensional profiles. Methods Enzymol. 1997;277:396-404. doi:10.1016/s0076- 6879(97)77022-8 38. Colovos C, Yeates TO. Verification of protein structures: patterns of nonbonded atomic interactions. Protein Sci Publ Protein Soc. 1993;2(9):1511-1519. doi:10.1002/pro.5560020916 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint 39. Wiederstein M, Sippl MJ. ProSA-web: interactive web service for the recognition of errors in three-dimensional structures of proteins. Nucleic Acids Res. 2007;35(Web Server issue):W407-410. doi:10.1093/nar/gkm290 40. Studer G, Rempfer C, Waterhouse AM, Gumienny R, Haas J, Schwede T. QMEANDisCo- distance constraints applied on model quality estimation. Bioinforma Oxf Engl. 2020;36(6):1765-1771. doi:10.1093/bioinformatics/btz828 41. Rodrigues CHM, Pires DEV, Ascher DB. DynaMut2: Assessing changes in stability and flexibility upon single and multiple point missense mutations. Protein Sci Publ Protein Soc. 2021;30(1):60-69. doi:10.1002/pro.3942 42. Laimer J, Hiebl-Flach J, Lengauer D, Lackner P. MAESTROweb: a web server for structure-based protein stability prediction. Bioinforma Oxf Engl. 2016;32(9):1414-1416. doi:10.1093/bioinformatics/btv769 43. Pires DEV, Ascher DB, Blundell TL. mCSM: predicting the effects of mutations in proteins using graph-based signatures. Bioinformatics. 2014;30(3):335-342. doi:10.1093/bioinformatics/btt691 44. Chen Y, Lu H, Zhang N, Zhu Z, Wang S, Li M. PremPS: Predicting the impact of missense mutations on protein stability. PLoS Comput Biol. 2020;16(12):e1008543. doi:10.1371/journal.pcbi.1008543 45. Parthiban V, Gromiha MM, Schomburg D. CUPSAT: prediction of protein stability upon point mutations. Nucleic Acids Res. 2006;34(Web Server issue):W239-242. doi:10.1093/nar/gkl190 46. Pancotti C, Benevenuta S, Birolo G, et al. Predicting protein stability changes upon single- point mutation: a thorough comparison of the available tools on a new dataset. Brief Bioinform. Published online January 11, 2022:bbab555. doi:10.1093/bib/bbab555 47. Lin YF, Cheng CW, Shih CS, Hwang JK, Yu CS, Lu CH. MIB: Metal Ion-Binding Site Prediction and Docking Server. J Chem Inf Model. 2016;56(12):2287-2291. doi:10.1021/acs.jcim.6b00407 48. Goodyear RJ, Richardson GP. Structure, Function, and Development of the Tectorial Membrane: An Extracellular Matrix Essential for Hearing. Curr Top Dev Biol. 2018;130:217-244. doi:10.1016/bs.ctdb.2018.02.006 49. Schraders M, Ruiz-Palmero L, Kalay E, et al. Mutations of the Gene Encoding Otogelin Are a Cause of Autosomal-Recessive Nonsyndromic Moderate Hearing Impairment. Am J Hum Genet. 2012;91(5):883-889. doi:10.1016/j.ajhg.2012.09.012 50. Yariz KO, Duman D, Zazo Seco C, et al. Mutations in OTOGL, encoding the inner ear protein otogelin-like, cause moderate sensorineural hearing loss. Am J Hum Genet. 2012;91(5):872-882. doi:10.1016/j.ajhg.2012.09.011 51. Zheng J, Miller KK, Yang T, et al. Carcinoembryonic antigen-related cell adhesion molecule 16 interacts with alpha-tectorin and is mutated in autosomal dominant hearing . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint loss (DFNA4). Proc Natl Acad Sci U S A. 2011;108(10):4218-4223. doi:10.1073/pnas.1005842108 52. Booth KT, Kahrizi K, Najmabadi H, Azaiez H, Smith RJ. Old gene, new phenotype: splice- altering variants in CEACAM16 cause recessive non-syndromic hearing impairment. J Med Genet. 2018;55(8):555-560. doi:10.1136/jmedgenet-2018-105349 53. Stooke-Vaughan GA, Obholzer ND, Baxendale S, Megason SG, Whitfield TT. Otolith tethering in the zebrafish otic vesicle requires Otogelin and α-Tectorin. Dev Camb Engl. 2015;142(6):1137-1145. doi:10.1242/dev.116632 54. Morgan CP, Krey JF, Grati M, et al. PDZD7-MYO7A complex identified in enriched stereocilia membranes. eLife. 2016;5:e18312. doi:10.7554/eLife.18312 55. Kammerer R, Rüttiger L, Riesenberg R, et al. Loss of mammal-specific tectorial membrane component carcinoembryonic antigen cell adhesion molecule 16 (CEACAM16) leads to hearing impairment at low and high frequencies. J Biol Chem. 2012;287(26):21584-21598. doi:10.1074/jbc.M111.320481 56. Cheatham MA, Goodyear RJ, Homma K, et al. Loss of the tectorial membrane protein CEACAM16 enhances spontaneous, stimulus-frequency, and transiently evoked otoacoustic emissions. J Neurosci Off J Soc Neurosci. 2014;34(31):10325-10338. doi:10.1523/JNEUROSCI.1256-14.2014 57. Xia A, Gao SS, Yuan T, et al. Deficient forward transduction and enhanced reverse transduction in the alpha tectorin C1509G human hearing loss mutation. Dis Model Mech. 2010;3(3-4):209-223. doi:10.1242/dmm.004135 58. Legan PK, Goodyear RJ, Morín M, et al. Three deaf mice: mouse models for TECTA- based human hereditary deafness reveal domain-specific structural phenotypes in the tectorial membrane. Hum Mol Genet. 2014;23(10):2551-2568. doi:10.1093/hmg/ddt646 59. Richardson G, Lukashkin A, Russell I. The tectorial membrane: One slice of a complex cochlear sandwich. Curr Opin Otolaryngol Head Neck Surg. 2008;16(5):458-464. doi:10.1097/MOO.0b013e32830e20c4 60. Alloisio N, Morlé L, Bozon M, et al. Mutation in the zonadhesin-like domain of alpha- tectorin associated with autosomal dominant non-syndromic hearing loss. Eur J Hum Genet EJHG. 1999;7(2):255-258. doi:10.1038/sj.ejhg.5200273 61. Huang RH, Wang Y, Roth R, et al. Assembly of Weibel–Palade body-like tubules from N- terminal domains of von Willebrand factor. Proc Natl Acad Sci U S A. 2008;105(2):482- 487. doi:10.1073/pnas.0710079105 62. Yasukawa R, Moteki H, Nishio S ya, et al. The Prevalence and Clinical Characteristics of TECTA-Associated Autosomal Dominant Hearing Loss. Genes. 2019;10(10):744. doi:10.3390/genes10100744 63. Kim DK, Kim JA, Park J, Niazi A, Almishaal A, Park S. The release of surface-anchored α- tectorin, an apical extracellular matrix protein, mediates tectorial membrane organization. Sci Adv. 2019;5(11):eaay6300. doi:10.1126/sciadv.aay6300 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint 64. Nam GS, Rim JH, Choi JY, et al. The TECTA mutation R1890C is identified as one of the causes of genetic hearing loss: a case report. BMC Med Genet. 2019;20(1):57. doi:10.1186/s12881-019-0775-1 65. Li Z, Guo Y, Lu Y, et al. Identification of a Novel TECTA mutation in a Chinese DFNA8/12 family with prelingual progressive sensorineural hearing impairment. PloS One. 2013;8(7):e70134. doi:10.1371/journal.pone.0070134 66. Ishikawa K, Naito T, Nishio SY, et al. A Japanese family showing high-frequency hearing loss with KCNQ4 and TECTA mutations. Acta Otolaryngol (Stockh). 2014;134(6):557-563. doi:10.3109/00016489.2014.890740 67. Rau A, Legan PK, Richardson GP. Tectorin mRNA expression is spatially and temporally restricted during mouse inner ear development. J Comp Neurol. 1999;405(2):271-280. 68. Calzada AP, Lopez IA, Ishiyama G, Ishiyama A. Otolithic Membrane Damage in Patients with Endolymphatic Hydrops and Drop Attacks. Otol Neurotol Off Publ Am Otol Soc Am Neurotol Soc Eur Acad Otol Neurotol. 2012;33(9):1593-1598. doi:10.1097/MAO.0b013e318271c48b . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint TABLES Table 1: Rare variants in the TECTA gene found in the studied familial MD cohort (N=99).

sequence: NM_005422.2 (TECTA); MAF: Minor allele frequency; NFE: Non-Finish European; MAX: Highest MAF value in gnomAD populations; ACMG: American College of Medical Genetics and Genomics; LP: Likely pathogenic; VUS: Variant of unknown significance Location Protein change Info Exon Domain MD family gnomADNFE MAF gnomADMAX MAF ACMG 11:121158016T>C p.Val1494Ala Missense 14 VWFD 4 F1 & F2 8.8x10-5 9.7x10-5 (AFR) VUS 11:121152980G>C p.Cys1402Ser Missense 13 TIL F3 1.5x10-5 4.8x10-4 (OTH) VUS 11:121157956AC>A p.Asn1474LysfsTer91 Deletion 14 - F4 Novel Novel LP 11:121165368C>T p.Pro1790Ser Missense 17 - F5 Novel Novel VUS 11:121189864GC>C p.Gly2118ProfsTer22 Deletion 23 - F6 0 2.4x10-5 (AFR) LP . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint FIGURE LEGENDS Figure 1: Variants distribution across the α-tectorin domains. The α-tectorin protein has several domains: the NIDO domain, the von Willebrand factor type C (VWFC: colored in green) and type D (VWFD) domains, the trypsin Inhibitor like cysteine rich (TIL) domain and the zona- pellucida (ZP) domains. Three rare missense variants (colored in blue) and two short deletions (colored in red) were found in four multicase MD families (arrow-shaped) and two additional families with one MD patient and relatives with partial syndromes (diamond-shaped). The position of stop codons generated by frameshift short deletions is indicated in red. Figure 2: Serial pure tone audiograms for MD patients from the six families included in this study carrying variants in the TECTA gene. The different lines in the audiograms show the hearing loss progression across the years (yo). Years after onset are indicated in brackets for each individual. Figure 3: Pedigrees of the six families carrying rare variants and short deletions in the TECTA gene. Figure 4: Predicted effect of variants on the α-tectorin structure. A) Mature α-tectorin model showing the different domains that form the protein and the positions of the missense variants found in this study (indicated with black arrows). B) TIL-3 domain of WT α-tectorin (left) and C1402S mutant (right). A disulfide bridge between C1359 and C1402 is predicted in the WT protein (indicated with an arrow). This disulfide bridge disappears in the mutated protein, where appears a weak polar interaction between C1359 and S1402. C) VWF D-4 domain of WT α- tectorin (left) and V1494A mutant (right). D) Interdomain upstream the ZP domain of α-tectorin WT (left) and P1790S mutant (right). Yellow dashed lines represent polar interactions between residues within 4 Å from the mutated residue. Arrows show new or missing bonds. VWF D: Von Willebrand factor type D; ZP: Zona Pellucida. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Figure 1 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Figure 2 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Figure 3 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Figure 4 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint SUPPLEMENTARY MATERIAL Ultrarare missense and frameshift variants in the TECTA gene may involve tectorial membrane in familial Meniere disease MEDRXIV PABLO ROMAN-NARANJO, ALBERTO M. PARRA-PEREZ, ALBA ESCALERA-BALSERA, ANDRES SOTO-VARELA, ALVARO GALLEGO- MARTINEZ, ISMAEL ARAN, NICOLAS PEREZ-FERNANDEZ, DAVID BÄCHINGER, ANDREAS H. ECKHARD, ROCIO GONZALEZ-AGUADO, LIDIA FREJO, JOSE ANTONIO LOPEZ-ESCAMEZ . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint SUPPLEMENTARY TABLES Supplementary Table 1 : Pair of primers used to validate candidate single nucleotide variants by Sanger Sequencing in the TECTA gene. Variant (hg38) Forward primer Reverse primer chr11:121152980G>C TCACCTGCCCTCCAAACAG CTCGTAATATTTGCCATCGGAGT chr11:121165368C>T CCATCTGACCATTTCCAATGTGA AAAAGGATGTAGCTGTACTTTGAAG chr11:121158016T>C ATGCAAGTCAGACGAGGAGT AAGGAGATGTCGGGCAGTTT . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Supplementary Table 2: Summary of the clinical information of familial MD patients carrying variants in the TECTA gene. Families F1 F2 F3 F4 F5 F6 Genetic variants TECTA gene p.Val1494Ala p.Val1494Ala p.Cys1402Ser p.Asn1474LysfsTer91 p.Pro1790Ser p.Gly2118ProfsTer22 MD patients Index III-11 III-6 II-3 II-1 II-3 III-2 Other II-6 III-7 III-2 I-2 I-2 - - Relatives with incomplete phenotype Episodic vertigo II-1, II-8, III-13, III-15, III- 19 III-8 - - III-1 I-2 Hearing loss II-3 II-6, III-8 - - - I-2, II-7 Clinical data of index patient Sex Female Female Female Female Male Female Laterality (ear) Unilateral (Right) Bilateral Unilateral (Right) Bilateral Bilateral Unilateral (Left) Age of onset 41-45 51-55 41-45 21-25 31-25 26-30 Autoimmune diseases No No No No No No . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Supplementary Table 3: Predicted structural model evaluation of each of the α-tectorin domains, the 4-, 8-, and 10 -domain assemblies of the protein using the DEMO method and the α -tectorin structural model predicted by Alphafold2 (located at https://alphafold.ebi.ac.uk/entry/O75443). Molprobity Score, Verify3D, ERRAT, ProSA-web and QMEANDisCo metrics were used in the evaluation. Molprobity Score is a weighted logarithmic combination of different geometric scores such as c lashscore, percentage of unfavoured Ramachandran and percentage of bad sidechain rotamers, giving a number that reflects the crystallographic resolution at which those values would be expected. Lower values of Molprobity Score are better. Verify3D determines the compatibility of an atomic model (3D) with its own amino acid sequence (1D) by assigning a structural type based on its location and environment. A higher score indicated high-quality of the structure. The overall quality factor, ERRAT, analyses the statistics of interactions between the different types of atoms and plots the value of the error function calculated by a comparison with highly refined structure statistics. As the generally accepted range for a high - quality model is >50, this analysis revealed that the backbone conformation and nonbonded interactions of all models were within the scope of a high-quality model. In the ProSA-web tool, the score is z -score defined as the energy separation between the native fold and the average of an ensemble of the misfolds in standard deviation units of the database. A z-score outside a range characteristic for native proteins of similar sizes indicated an erroneous structure. In this case, each model is in the range. Finally, QMEANDisCo evaluates the ag reement of pairwise distances between residues with sets of distance constraints extracted from structures homologous to the evaluated model, so the higher the score the better the model. * Selected structural model domain to build the assembly and model the whole protein. ** Model used for evaluating the impact of the variants found in this study. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Protein Domain Initial domain position Final domain position Initial modelled domain position Final modelled domain position Modelled length Modelling method Template used Evaluation Molprobity Score Verify3D ERRAT ProSA-Web QMEANDisCo NIDO 98 252 23 252 230 Modeller No template available - - - - - Robetta-AB - 1.79 85.81 56.0606 -5.57 0.47 CI-TASSER - 4.19 98.06 46.25 -5.29 0.34 AlphaFold2 * - 1.13 100 87.7729 -5.86 0.68 VWFC 260 314 253 314 62 Modeller No template available - - - - - Robetta-AB * - 0.83 96.36 82.6087 -4.94 0.46 CI-TASSER - 2.85 38.18 91.4894 -5.16 0.46 AlphaFold2 - 1.74 71.67 65.1163 -4.65 0.47 VWFD 1 320 500 315 500 186 Modeller 7KWO_V 3.2 59.55 47.6415 -3.66 0.55 AlphaFold2 * - 1.39 93.84 94.697 -6.36 0.55 TIL 1 597 650 501 650 150 Modeller 7A5O_E 2.8 52.23 52.22 -2.87 0.47 AlphaFold2 * - 0.5 71.05 92.1348 -4.79 0.63 VWFD 2 711 886 651 886 236 Modeller 7KWO_V 3.38 80.18 21.1538 -5.2 0.55 AlphaFold2 * - 1.04 99.63 95.2 -6.02 0.6 TIL 2 984 1036 887 1036 150 Modeller 7A5O_E 2.8 52.83 52.2727 -2.76 0.46 AlphaFold2 * - 0.5 75.44 100 -6.11 0.59 VWFD 3 1098 1278 1037 1278 242 Modeller 7KWO_V 3.52 72.6 29.8578 -4.64 0.54 AlphaFold2 * - 0.83 97.45 96.3415 -7.33 0.65 TIL 3 1372 1425 1279 1425 147 Modeller 7A5O_E 3.12 37.04 13.6364 -2.93 0.49 AlphaFold2 * - 0.94 81.42 100 -5.16 0.59 VWFD 4 1485 1666 1426 1666 241 Modeller 7KWO_V, 7A5O_E 3.46 95.38 43.6364 -5.58 0.61 AlphaFold2 * - 1.18 93.96 89.441 -6.27 0.67 ZP 1805 2059 1667 2059 393 Modeller 6ZS5, 4WRN, 6TQL 3.49 73.73 36.3265 -5.31 0.58 AlphaFold2 * - 1.61 65.88 94 -9.1 0.53 Complete α- tectorin (O75443) 1 2155 1 2155 2155 AlphaFold2 - 1.52 - 89.5459 -17.6 0.57 4 Domains Assembly 98 650 23 650 628 AlphaFold2 - 1.47 90.7 83.7879 -9.1 0.6 8 Domains Assembly 98 1425 23 1425 1403 AlphaFold2 - 1.13 - 82.1324 -14.93 0.55 10 Domains Assembly ** 98 2059 98 2059 1962 AlphaFold2 - 1.05 - 80.6365 -17.29 0.55 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Supplementary Table 4: Probable Ca2+ binding sites in the α-tectorin VWFD4 protein structure. Binding site α-tectorin residue number Amino acid Score (MIB) 1 1507 ALA 2.096 1508 ASN 2.096 2 1530 ASN 1.682 1531 PHE 1.682 1532 ASP 1.682 3 1624 ASN 2.143 1626 ASN 2.143 1627 GLY 2.099 1628 ASP 2.143 1630 THR 1.641 1631 ASP 2.143 1632 ASP 2.143 1499 ASP 1.746 4 1664 SER 2.097 1666 ASN 2.097 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Supplementary Table 5: Predicted α-tectorin stability perturbation caused by missense variants found in this study. Predictor Variants p.Cys1402Ser p.Val1494Ala p.Pro1790Ser Maestro-web (kcal/mol) Neutral (0.110) Neutral (0.056) Neutral (0.022) mCSM (stability) (kcal/mol) Destabilizing (-1.254) Destabilizing (-1.099) Destabilizing (-0.545) CUPSAT thermal (kcal/mol) Destabilizing (-13.02) Destabilizing (-0.57) Stabilizing (2.23) CUPSAT denaturants (kcal/mol) Stabilizing (12.46) Stabilizing (0.79) Neutral (-0.4) DynaMut2 (kcal/mol) Destabilizing (-0.54) Destabilizing (-1.29) Neutral (-0.25) PremPS (kcal/mol) Destabilizing (2.2) Destabilizing (0.6) Neutral (0.43) Global protein stability change prediction (kcal/mol) in the α-tectorin model using different ΔΔGpred prediction methods. For Maestro-web and PremPS, ΔΔGpred 0.0 indicates a stabilizing mutation. Variants were considered neutral in terms of protein stability perturbation when -0.5 < ΔΔGpred < 0.5. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint SUPPLEMENTARY FIGURES Supplementary Figure 1: Validation of candidate variants by Sanger sequencing. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint Supplementary Figure 2: The four predicted Ca2+ binding sites in the α-tectorin VWFD4 protein structure (Supplementary Table 3 – coloured in this figure in green) located near the variant p.Val1494Ala , found in F1, F2 and a sporadic patient with MD. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted March 3, 2022. ; https://doi.org/10.1101/2022.02.18.22270926doi: medRxiv preprint