Haploinsufficiency of KPNA7 causes otosclerosis, likely due to the release of import inhibition of PTHrP and the reactivation of chondrogenesis in the globuli interossei

19 20 Otosclerosis is a genetic bone disorder restricted to the otic capsule and a common cause of 21 conductive hearing loss with both familial and sporadic cases. To date, 14 genomic loci (OTSC) 22 and four underlying OTSC genes (MEPE, SERPINF1, FOXL1, SMARCA4) have been identified 23 in autosomal dominant families. A combined genetic/genomics approach on five affected 24 siblings of Northern European ancestry from the island of Newfoundland, Canada identified a 25 premature stop mutation in Karyopherin subunit α7 (KPNA7, c.49C>T, p.R17X). KPNA7 maps 26 to OTSC2 (7q22.1) and encodes the newest of the seven-member importin-α family of nuclear 27 transporters and plays a critical role in early embryonic cleavage events and zygotic genome 28 activation. Previous studies reveal that recessive KPNA7 variants cause skeletal abnormalities, 29 including scoliosis and ocular hypertelorism in two sisters with Partial Corpus Callosum 30 Agenesis-Cerebellar Vermis Hypoplasia With Posterior Fosa Cysts Syndrome and more 31 recently, have been implicated in preimplantation embryo arrest (PREMBA) (OMIM 614107). 32 Interestingly, KPNA7 is also a maternal factor with an exclusively embryonic role and likely 33 inhibits non-classical NLS transport of PTHrP, a known activator of chondrogenesis. We 34 propose that KPNA7 haploinsufficiency causes a failure in nuclear transport inhibition of PTHrP 35 in the quiescent embryonic cells of the globuli interossei in the otic capsule and re-activates 36 chondrogenesis. The KPNA7 discovery provides new insights into the pathogenesis of 37 otosclerosis and potential for targeted therapies. 38 39 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 3 Author Summary 40 Otosclerosis is a distinctly human genetic bone disorder of the otic capsule and a major cause of 41 progressive hearing loss in young adults, particularly in females. Even though otosclerosis has 42 been recognized as a distinct entity for a long time, both its pathogenesis and restriction to the 43 otic capsule remains a mystery. Here, we use a combined genetic/genomics approach to identify 44 a premature stop mutation in five affected siblings of Northern European ancestry from the 45 island of Newfoundland, Canada. KPNA7 encodes the newest of the seven-member importin-α 46 family of nuclear transporters and plays a critical role in early embryonic cleavage events and 47 zygotic genome activation. Based on the unique features of the otic capsule, we hypothesize that 48 the premature stop mutation in KPNA7 leads to haploinsufficiency causing a failure in nuclear 49 transport inhibition of PTHrP and reactivates chondrogenesis in the otherwise quiescent 50 embryonic cells within the otic capsule. The KPNA7 discovery provides new insights into the 51 pathogenesis of otosclerosis and potential for targeted therapies. 52

53 54 Otosclerosis is a uniquely human skeletal disorder restricted to the otic capsule of the temporal 55 bone and a common cause of progressive conductive hearing loss (HL) in young adults. An 56 autosomal dominant (AD) disease with environmental triggers, otosclerosis is clinically 57 characterized by abnormal bone deposition in the middle ear, distorting the fine structures of the 58 ossicular chain and limiting the movement of the stapes bone against the oval window. The 59 immobilization of the stapes bone results in conductive HL as well as sensorineural HL in some 60 patients due to abnormal bone growth extending into the fluid-filled inner ear (1). The location 61 of the inner ear within the osseous labyrinth greatly impacts perilymph sampling for diagnostic 62 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 4 purposes and local drug delivery. Visualization during surgery to replace or repair stapes fixation 63 due to otosclerosis validates the diagnosis, and in most cases, restores the conductive component 64 of hearing to those able to access stapes surgical prosthesis replacement. Otosclerosis risk factors 65 include positive family history, sex (female), measles and pregnancy (1, 2). Recognised as a 66 medical entity some 125 years ago (3), this restricted bony disorder cannot be predicted, stopped 67 or medically treated, and its pathogenesis remains a mystery. 68 69 Despite several decades of research efforts, the otosclerosis (OTSC) genes have been recalcitrant 70 to discovery because of the genetically heterogeneous nature of otosclerosis and the rarity of AD 71 families under study (2). So far, 14 distinct (OTSC1-14) loci have been mapped in AD families 72 and four causative OTSC genes (MEPE, SERPINF1, FOXL1, SMARCA4) identified (Hereditary 73 Hearing Loss Homepage). A recent search for susceptibility factors involving 3504 otosclerosis 74 cases from three biobank studies revealed 23 novel loci linked to genes whose dysregulation in 75 bone remodeling and mineralization causes rare monogenic skeletal disorders (4). These near 76 protein associations provide insight into the nature of the highly penetrant OTSC genes but not 77 their identity, as GWAS studies exclude rare variants by design. Given the high genetic 78 heterogeneity underlying skeletal dysplasias, with 461 genes known to cause monogenic forms, 79 more OTSC genes are anticipated (5). 80 81 The otic capsule has a highly complex anatomy, and its embryonic development is one of the 82 most complicated examples of cellular morphogenesis in any biologic system (6). In temporal 83 bone, the inner ear tissues and spaces are enclosed within the bony otic capsule, the hardest bone 84 in the body, a critical feature essential to maintaining hearing integrity (7). The otic capsule 85 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 5 forms through endochondral ossification, one of two essential pathways of bone formation that 86 uses cartilage as a bone template during fetal development. Mesenchymal stem cells differentiate 87 into chondrocytes (cartilage cells) which proliferate rapidly, hypertrophy and secrete the 88 extracellular matrix that undergoes mineralization. Eventually the hypertrophic chondrocytes die 89 through apoptosis and are replaced by osteocytes that become trapped in bony matrix. Although 90 the otic capsule is fully formed by the fifth fetal month (8), islands of embryonic tissue 91 containing quiescent chondrocytes and osteocytes, known as the globuli interossei, are uniquely 92 retained by the otic capsule and persist throughout life. These embryonic remnants are 93 increasingly implicated as the site of otosclerosis in the temporal bone (9-11). 94 95 The human skeleton continues to grow and repair (remodel) postnatally via the well-studied 96 RANK/RANKL/OPG pathway. In contrast, the otic capsule and the ossicular chain are fully 97 formed in utero and remodeling is virtually absent due to the overproduction of OPG 98 (osteoprotegerin) (12). In the bony lacunae of the otic capsule, osteocytes communicate via an 99 intercellular canalicular network that provides nutrient transport and bathes the perilacunar 100 matrix with OPG. As a normal part of aging, osteocytes die but are not replaced as OPG prevents 101 osteoclasts from maturing and beginning the remodeling process (13). How bone remodeling in 102 the otic capsule occurs in the presence of overproduction of OPG close to the inner ear is a 103 mystery, and suggests other parallel protective mechanisms must play a role (13, 14). 104 105 Herein we identify a stop mutation in KPNA7 (Karyopherin subunit α7), the newest member of 106 the Importin-α (Imp-α) transport factors involved in nucleocytoplasmic trafficking, in a white 107 family of Northern European descent. Nucleocytoplasmic trafficking is a highly efficient and 108 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 6 regulated system comprising of 60 proteins of the nuclear transport system where dysregulation 109 is linked to major diseases such as cancer, viral infections, inflammation and neurodegenerative 110 diseases, making them prime targets for therapies (15). Expression of KPNA7 is highly restricted 111 to oocytes and early embryogenesis where recessive mutations cause congenital skeletal 112 abnormalities (16) and preimplantation embryo arrest (PREMBA) (17). Perhaps most significant 113 to otosclerosis, Imp-αs are known to inhibit the nuclear import of parathyroid related protein 114 (PTHrP), a major regulator of chondrogenesis (18). We explore how embryonically expressed 115 genes such as KPNA7 may cause adult-onset HL and garner insights into the pathobiology of 116 otosclerosis and why the rest of the human skeleton is likely spared from abnormal bone 117 deposition and disease. 118

119 Clinical recruitment, pedigree structure and classification of hearing loss 120 The proband (PID III-1) was diagnosed at 42 with severe HL due to otosclerosis in both ears, 121 which began in teenage years. Pre-surgery audiogram revealed bilateral, conductive HL with 122 borderline cochlear (sensorineural) loss. Hearing improved bilaterally after two successive 123 stapedectomies (Fig 1). The diagnosis of otosclerosis was confirmed upon surgical visualization 124 of stapes fixation. Based on the medical questionnaires, HL started in the teens for all but one 125 sibling who noticed HL in adulthood, and hearing was restored in all siblings after stapedectomy 126 surgery. The pedigree structure is consistent with both AD and autosomal recessive (AR) 127 inheritance, but X-linked inheritance can be ruled out as otosclerosis is not more severe in males. 128 Haplotype sharing on chromosome 7q and targeted gene sequencing 129 When testing for linkage to mapped OTSC loci and recapitulating disease-associated haplotypes, 130 we would expect, under an AD model, to observe a single otosclerosis-associated haplotype that 131 is shared exclusively among affected family members. Conversely, under an AR model, we 132 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 7 would expect all affected to share the same maternal and paternal disease haplotypes. Although 133 we have limited clinical data on the paternal side, under the AD model, we observed a shared 134 paternal disease haplotype encompassing OTSC2 (7q) (Fig 2). Conversely, no sharing was 135 observed for OTSC1 (15q), OTSC3 (6p), OTSC4 (16q), OTSC5 (3q), OTSC7 (6q), OTSC8 (9p) 136 or the COL1A1 (17q) and NOG (17q) genomic regions. Subsequent targeted gene sequencing of 137 11 positional candidate genes within the disease interval on 7q did not identify the otosclerosis 138 gene. The proband also screened negative for the 15-base pair (bp) deletion in FOXL1 139 (rs764026385; OTSC11) that we previously identified in a Newfoundland family, and for rare 140 otosclerosis variants in SERPINF1. 141 Genome wide analysis under AR linked regions 142 Although there was evidence for haplotype sharing on 7q, targeted gene sequencing did not 143 identify the gene. Under an AR model, linkage simulation derived the theoretical maximum 144 LOD score (LODmax=2.5) given the pedigree structure. SNP genotyping and multipoint linkage 145 analysis yielded LODobs=2.5 at chr17q25.1-q25.3, spanning a region of 5.9 Mb. Exome 146 sequencing of the 141 positional candidate genes revealed zero homozygous variants and 49 147 heterozygous variants. Only two genes, TEN1 and EVPL, had two or more variants consistent 148 with AR inheritance however, these were filtered out due to high population frequencies and/or 149 benign functional predictions (19). 150 Genotype wide analysis under AD linked regions 151 Assuming AD inheritance, the LODmax (1.73) was not obtained but suggestive LOD (LOD > 1) 152 scores were observed at five distinct genomic loci (7q, 10p, 10q, 16q, 17q). Several of these 153 overlapped with OTSC disease intervals, specifically OTSC2 (7q), OTSC4 (16q), NOG ( 17q). 154 We identified 542 positional candidate genes under the five suggestive loci. Exome sequencing 155 yielded 301 variants and of these,153 variants were filtered out as they were not identified in all 156 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 8 affected. We could remove 37 variants as they were also identified in "solved" FOXL1 cases. Of 157 the remaining variants, 69 had MAF >2%, reducing the variants of interest to 23 silent, 15 158 missense, one intronic and one nonsense. The 15 missense mutations were all predicted to be 159 benign, and 16 (15 silent, one intronic) were predicted to have no effect on splicing. Only the 160 nonsense mutation in KPNA7 gene on 7q (OTSC2) remained after variant filtering. 161 Sanger validation, cascade sequencing and in silico analyses 162 KPNA7, c.49 C>T (NM_001145715.3) is rare (0.015%, gnomAD, rs746784660) and has been 163 reported in ClinVar as a VUS (ID: 652650). We found that KPNA7, c.49 C>T is absent in 187 164 HL probands and in controls. Cascade sequencing confirmed co-segregated with otosclerosis 165 (Fig 2). KPNA7, c.49 C>T (NM_001145715.3) is located near the 5' end of KPNA7 and is 166 predicted to cause premature stop mutation, p. R17X, subjecting the truncated RNA transcript to 167 nonsense mediated decay and resulting in haploinsufficiency (Alamut Visual Plus, version 1.13, 168 Sophia Genetics (2024)). Previously, recessive mutations in KPNA7 cause congenital skeletal 169 abnormalities (16) and preimplantation embryo arrest (PREMBA) (17). We conclude that 170 KPNA7, c.49 C>T, p. R17X is pathogenic according to HL ACMG criteria PVS1 and PM2 (20, 171 21). 172

173 174 Summary of findings 175 We identify KPNA7, the newest of the human Imp-α transport factors, as the first of the nuclear 176 transport system of proteins to cause AD otosclerosis. Surgical reports in a white multiplex 177 family of Northern European extraction confirmed the diagnosis and affected siblings reported 178 HL as young adults. Using comprehensive genetic and genomics analyses, we identified a 179 disease-associated haplotype on chromosome 7q22.1 overlapping the OTSC2 locus (22). The 180 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 9 absence of paternal DNA was overcome by recruiting maternal siblings, confirming paternal 181 transmission of a premature stop mutation in KPNA7 [(NM_001145715.3), c.49 C>T, p. R17X]. 182 KPNA7 is located within the OTSC2 locus and resides in the vicinity of RELN (OMIM 605727), 183 a gene whose intronic variants have been validated in GWAS otosclerosis case studies (4, 23). 184 Although not normally expressed in adult tissues, KPNA7 is reactivated in cancer cell lines, with 185 the highest expression detected in pancreatic cell lines harbouring an amplification of the 7q21-186 22 genomic locus where KPNA7 resides (24). 187 Otosclerosis may be the failure of two or more parallel protective mechanisms 188 Bloch and colleagues have eloquently modeled how age-dependent microdamage accumulates in 189 the human perilabyrinthine bone where bone remodeling is essentially absent (13, 25-29). 190 Clusters of dead osteocytes (cellular voids) lose connection with each other as the perilacunar 191 matrix breaks down, rendering OPG and other nutrients beyond the reach of viable osteocytes 192 trapped within these voids (26). In their model of aging, quiescent cells within the globuli 193 interossei may be able to break free of suppressive actions of OPG and go on to complete the 194 process of endochondral ossification. However, the accumulation of voids alone is unlikely to be 195 causative as clinical otosclerosis is not prevalent (0.30–0.38% in Europeans) and histologic 196 otosclerosis has been observed in under 3% of temporal bone autopsy specimens (30). Tissue 197 from patients with labyrinthitis ossificans, a rare reaction to infection where the membranous 198 labyrinth turns to bone, essentially halting OPG production, show the persistence of cartilage 199 remnants and the absence of bone remodeling, suggesting parallel mechanisms, in addition to 200 OPG, protect the otic capsule from remodeling (27). Our exploration into the function of 201 KPNA7 in embryonic cells suggests that KPNA7 represents a parallel protective mechanism. 202 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 10 The structure of Imp-αs and their many roles in nuclear transport 203 Cells require an active transport machinery to shuttle RNA, protein and other macromolecules to 204 their correct subcellular localizations to maintain homeostasis and carry out normal functions. 205 For proteins, the most utilized nuclear import pathway is mediated by Imp-αs (31). KPNA7 206 encodes the newest of the seven-member imp-α karyopherins, a group of highly conserved 207 proteins that share a common structure consisting of a body of ten helical Armadillo (ARM) 208 repeats, a short C-terminal region of acidic amino acids and an N-terminal, Importin- β (Imp-β) 209 binding (IBB) domain. Most proteins targeted for transport into the nucleus contain a nuclear 210 localization signal (NLS) motif, a lysine-rich stretch of basic amino acids containing one 211 (monopartite) or two (bipartite) basic regions separated by a linker region in their amino acid 212 sequence (31). The first NLS motif to be recognized and best characterized is the classical NLS 213 (cNLS). Imp-αs function as adaptors that recognize and bind to cNLS in their cargo proteins. 214 Proteins with non-classical NLS (ncNLS) motifs can also be bound and transported directly by 215 Imp-βs without the need for Imp-α adaptors. 216 217 Except for KPNA7, Imp-αs are maintained in a closed state (autoinhibited) in the cytoplasm with 218 the highly flexible IBB domain folding back onto itself blocking the NLS binding groove from 219 binding NLS-containing cargo (32, 33). In cNLS transport, the IBB domain binds Imp-β, 220 exposing the NLS binding groove for specific cargo protein binding. The cargo-Imp-α-β 221 tricomplex, once formed, is rapidly transported to the nucleus via Imp-β interactions with select 222 nucleoporins lining the central channel of the nuclear pore complex. In ncNLS transport, Imp-β 223 binds and transports NLS-containing cargo in the absence of Imp-α. Inside the nucleus, Imp-β 224 binds RanGTP and releases Imp-α and cargo protein and then Imp-α and Imp-β are recycled 225 back to the cytoplasm by export receptors. 226 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 11 Imp-αs can also act as negative regulators for the nuclear import of certain proteins by 227 competing with Imp-β for NLS binding to cargo, or by forming a transport-incompetent complex 228 in the cytoplasm, preventing cargo from entering the nucleus. For example, in the presence of 229 Imp-α, TRF1 forms a complex with Imp-α-β but this complex remains in the cytoplasm. In 230 contrast, Imp-α competes with the binding of imp-β to Snail zinc finger domain, resulting in 231 ineffective nuclear accumulation of Snail, leading to a decrease in its cellular protein level 232 through subsequent degradation by the protease system, with implications for the prevention of 233 tumor cell invasion by inhibiting Snail localization (34). Perhaps most significant to otosclerosis, 234 Imp-αs are known inhibitors of PTHrP, a multifunctional cytokine and a major regulator of 235 chondrogenesis in the human skeleton sharing structural similarities with, and the same receptor 236 (PTH1R) as parathyroid hormone (PTH) (35). Interestingly, PTH1R mRNA expression in 237 otosclerotic stapes led Grayeli and colleagues to hypothesize that abnormal cellular response to 238 PTH played a role in abnormal remodeling in otosclerosis (36). 239 Characteristics of KPNA7, the newest nuclear import factor 240 KPNA7 is the most recent and divergent of the seven human Imp-αs and an intriguing OTSC 241 gene as it is not normally expressed in adult tissues. KPNA7 is recognised as a maternal factor 242 essential for embryogenesis and fertility (37) and plays a critical role in protein transport in 243 oocytes and early embryos and is critical to early embryonic cleavage events and zygotic 244 genome activation (38). KPNA7 is known to have many cargo proteins, up to 377 have been 245 reported (24). KPNA7 has a unique ability among the Imp-αs to maintain an open state and has 246 the strongest IBB domain capacity for Imp-β, likely critical to producing pre-formed Imp-α-β 247 heterodimers to increase the transport rate of cargo proteins in the early stages of embryogenesis 248 (38). KPNA7 is the most abundant Imp-α in germinal vesicle and metaphase II-stage oocytes 249 (39) and is nearly absent from the eight-cell embryo onward (38), being rapidly degraded during 250 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 12 zygotic genome activation and barely detectable in morula- and blastocyst-stage embryos (40). 251 Recessive KPNA7 mutations result in low protein expression levels, interfering with the nuclear 252 import of RSL1D1 (also known as cellular senescence-inhibited gene protein: CSIG) (17). 253 RSL1D1 negatively regulates PTEN via translational suppression causing increased cell 254 proliferation, is significantly elevated in the tumors of colorectal cancer patients predicting 255 poorer survival outcomes and is a potential new target for cancer therapies (41). Imp-α has been 256 shown to inhibit ncNLS transport of certain proteins including PTHrP (42). As the N-terminal 257 domain of Imp-β binds PTHrP at HEAT repeats 2-11, but also binds Imp-α (IBB domain) at 258 HEAT repeats 7-19, this partial overlap of binding sites may explain ncNLS transport inhibition 259 of PTHrP (43). 260 Nuclear import of PTHrP drives developmental pathways in a context-specific manner 261 The otic capsule forms through a series of molecular and cellular signaling processes where 262 mesenchymal progenitor cells undergo condensation and differentiation into chondrocytes which 263 subsequently proliferate and hypertrophy, followed by mineralization of the extracellular matrix 264 and apoptosis of the chondrocytes (44) (Fig 3). PTHrP is necessary for endochondral 265 ossification, regulating chondrocyte maturation, proliferation and differentiation (44). In fact, 266 PTHrP is essential to development. PTHrP null mice (Pthrp -/-) die early in the postnatal period 267 and display severe chondrodysplasia with reduced endochondral development and excessive 268 mineralization (45). Experiments testing the effect of mechanical strain on chondrocytes showed 269 that PTHrP expression increased during the proliferation and matrix forming stages under 270 conditions of cyclical strain due to Indian Hedgehog (Ihh) signaling in chondrocytes (46). The 271 Ihh signaling pathway also works with other signaling molecules and pathways to promote 272 chondrocyte proliferation and inhibit hypertrophy and forms a negative feedback loop with 273 PTHrP pivotal in regulating cartilage development (47). Upon activation of Ihh expression from 274 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 13 an external trigger such as mechanical stress, Ihh binds to the transmembrane protein Patched 275 (Ptc), releasing inhibition of Smoothened (Smo) at the cell surface, which then activates the 276 expression of the downstream signaling molecule Gli. Gli enters the nucleus and regulates the 277 expression of downstream signaling factors SOX9, RUNX2 and PTHrP (48) (Fig 3). 278 How KPNA7 haploinsufficiency may cause otosclerosis 279 As PTHrP nuclear import is integral to its function in cells, strategies to block PTHrP nuclear 280 import could have important effects on target cell function (18). Endochondral ossification is 281 initiated by nuclear import of PTHrP and quiescent chondrocytes are known to express PTHrP 282 (49). Upon entering the nucleus by ncNLS transport chondrocyte regulation is achieved through 283 the Ihh-PTHrP pathway (44), as well as the WNT/ β-catenin pathway (50) via negative feedback 284 loops. PTHrP is both fast acting and fast to be removed (51). In a scenario where the KPNA7-285 Imp-β complex is pre-formed, potentially KPNA7 would be able to prevent binding of PTHrP by 286 binding a different cargo or outcompete PTHrP and enter the nucleus in a cargo-free state, 287 thereby inhibiting nuclear transport of PTHrP, as suggested by Oostdyk and colleagues for other 288 cargo proteins (38). While KPNA7 has a time-limited role in the nuclear import of factors 289 required for maternal-to-zygotic transition and early embryogenesis, it is logical to think that 290 KPNA7 also inhibits endochondral ossification by preventing nuclear import of PTHrP in the 291 unique and peculiar case of the embryonic cells of the globuli interossei. 292 293 Although not normally expressed in adult tissues, KPNA7 is reactivated in cancer cell lines 294 promoting carcinogenesis by increasing the rate of import of cell cycle factors due to genomic 295 amplification on 7q21-22. Unlike amplification in pancreatic cancer, the effect of a heterozygous 296 KPNA7 stop mutation identified in otosclerosis patients is predicted to cause its phenotypic 297 effects via haploinsufficiency. Therefore, it is unlikely that otosclerosis is due to increased 298 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 14 import of KPNA7 cargo proteins. The quiescent chondrocytes of the globuli interossei, unique to 299 the otic capsule, would provide a likely substrate for KPNA7 (10, 11). While a mutation in an 300 Imp-α exclusively expressed in early development would not be expected to impact the rest of 301 the human skeleton, we hypothesize that KPNA7 plays a key role in preventing PTHrP nuclear 302 import in quiescent chondrocytes. If this is indeed the case, KPNA7 haploinsufficiency may 303 activate chondrogenesis leading to the deposition of spongy, disorganized bone characteristic of 304 osteosclerotic specimens, and specifically, the mineralization stage of the extracellular matrix in 305 hypertrophic chondrocytes. If this speculation is true, otosclerosis is not the result of pathologic 306 bone resorption and deposition, but one of re-activation of chondrogenesis, providing an 307 alternative mechanism for its pathogenesis. 308 OTSC genes and their roles in endochondral ossification 309 A closer look at the OTSC genes MEPE, FOXL1 and SMARCA4 reveal they are involved in 310 endochondral ossification signaling pathways (Fig 3). In addition, the OTSC gene SERPINF1 311 encodes for PEDF, which is involved in many biological processes including bone formation, 312 binding to extracellular matrix proteins including collagen and glycosaminoglycan, and is 313 involved in the mineralization of bone matrix (52). Rämö and colleagues identified otosclerosis 314 susceptibility genes BMP-2,-3,-4,-7 (involved in bone development) and SOX9 and RUNX2 315 (4). MEPE normally serves as a decoy receptor for pre-osteoclasts inhibiting osteoclast 316 maturation and its ASARM motif, upon proteolytic cleavage by β-cathepsin, inhibits 317 mineralization by binding to hydroxyapatite crystals (53). In zebrafish, foxl1 (OTSC11) regulates 318 the expression of collagen genes such as col1a1 and col11a2, and results in a delay in jawbone 319 mineralization (54). SMARCA4 (also known as BRG1) plays a role in ossicle formation during 320 embryogenesis and may be important for regulation of osteoblast differentiation and 321 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 15 maintenance of postnatal bone homeostasis in the otic capsule (55). WNT/ β-catenin is also 322 involved in proliferation and differentiation through a negative feedback loop with PTHrP (50). 323 Re-activation of endochondral ossification pathways in the globuli interossei, a new 324 mechanistic model for otosclerosis 325 Otosclerosis is likely due to germline susceptibility factors interacting with environmental 326 triggers (measles infection, pregnancy, etc.) to generate the pathological phenotype (2). Perhaps 327 deficient KPNA7 causes the reactivation of chondrogenesis and the resumption of endochondral 328 ossification after loss of repression by OPG in aging otic capsules. From this perspective, the 329 pathologic spongy bone deposition seen in otosclerosis may be the result of dysregulation caused 330 by mutations in OTSC genes regulating endochondral ossification. Endochondral ossification 331 involves multiple signaling pathways including the Ihh pathway, which regulates chondrocyte 332 maturation and bone formation, the PTHrP pathways and bone morphogenetic proteins (BMPs). 333 Ihh regulates chondrocyte maturation and bone formation, and bone morphogenetic proteins 334 (BMPS) induce chondrocyte differentiation via regulating the expression of SOX9 and 335 stimulating endochondral ossification via transcriptional regulation of RUNX2. 336 Could dysregulation of PTHrP due to KPNA7 cause otosclerosis in women. 337 In a literature review on otosclerosis and pregnancy, Fabbris and colleagues found that the only 338 significant correlation was between pregnancy and disease onset, noting that cases with hearing 339 impairment increased with number of pregnancies (56). We know that PTHrP levels are higher 340 in pregnant women, and are significantly higher in lactating women, being detected in breast 341 milk at levels exceeding 10,000 times those found in the blood of hypercalcemia of malignancy 342 patients or normal controls (57). Otosclerosis during pregnancy may be the result of increased 343 bone remodeling (RANK/RANKL/OPG) (57-59) or due to susceptibility variants in KPNA7 or 344 other Imp-αs involved in regulating the actions of PTHrP. 345 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 16 Overlapping therapeutic targets in skeletal disorders and cancer 346 OTSC genes FOXL1, SMARCA4 and KPNA7 are increasingly recognised for their role in 347 carcinogenesis (60-64). By extension, therapeutic targets for cancer, otosclerosis and other 348 skeletal disorders increasingly overlap because the development, proliferation and migration of 349 cancer cells mimic critical pathways of embryonic development. This includes deregulation in 350 nucleocytoplasmic transport. For example, expression of KPNA7 in adult tissues is almost non-351 existent, except in pancreatic cancer cell lines where KPNA7 causes a significant decrease in cell 352 growth due to G1 arrest, accompanied by an increased expression of p21, a key regulator of the 353 cell cycle (65). As we have seen, low KPNA7 levels causes PREMBA through dysregulated 354 import of RSL1D1, which in turn, negatively regulates PTEN and is a potential new target for 355 colorectal cancer (41). Imp-αs also facilitate the nuclear import of Smad proteins in the TGF- 356 β/Smad3 pathway and show promise as therapeutic targets in rotator cuff injuries (66). There is 357 an epidemic of colorectal cancer in young people (67) and of adult-onset skeletal disorders 358 causing a public health issue in the global North. An investment in therapeutic options for cancer 359 and for the repair of cartilage for aging knees and hips may benefit patients with otosclerosis by 360 providing target drugs for off-label clinical trials. 361

of this study 362 Although KPNA7 maps to the OTSC2 locus (22), we did not have access to the family used to 363 map OTSC2, therefore we cannot confirm or deny that KPNA7 is OTSC2. Association between 364 otosclerosis and RELN may be due to the close physical proximity of RELN to disease-variants 365 in KPNA7. We assume equal expression of maternal and paternal KPNA7 alleles in the globuli 366 interossei however we do not have patient-derived disease tissue to examine the functional 367 consequences of KPNA7, c.49C>T, p.R17X. Otosclerosis due to KPNA7 haploinsufficiency may 368 be related to the dysregulated import of RSL1D1 or there may be alternative mechanistic 369 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 17 explanations for the pathogenesis of KPNA7 as the specific binding affinities and cargo 370 preferences of KPNA7, being the most recent member of the Imp-αs, have not been fully 371 explored. Evidence from this study suggests that reactivation of endochondral ossification, not the 372 de-repression of bone remodeling, underlies otosclerosis due to KPNA7 mutation. Future 373 research is warranted along this direction to describe the modulators of KPNA7 function and to 374 design potential therapies to modulate the PTHrP- Imp-α interactions for therapeutic purposes. 375 Future directions 376 We used strict clinical criteria and a combination of genetic and genomic analyses to 377 successfully identify a novel OTSC gene, KPNA7, the second OTSC gene identified in the white 378 settler population of the island of Newfoundland, Canada. KPNA7 is the first member of the 379 nucleocytoplasmic trafficking system to cause otosclerosis. We propose that normally, KPNA7 380 plays a critical role of inhibiting the nuclear import of PTHrP in the globuli interossei and that 381 KPNA7 haploinsufficiency results in reactivation of endochondral ossification, providing an 382 additional or alternative paradigm to the central hypothesis that dysregulation of bone 383 remodeling repression causes otosclerosis. As the molecular signaling pathways involved in 384 embryogenesis also play a role in the fate of neoplastic cells, investigating the shared cellular 385 and molecular signaling pathways and molecules that cause other monogenic skeletal disorders 386 and carcinomas may provide unique insights into novel therapeutic targets for common maladies 387 including otosclerosis, osteoarthritis and cancer. Difficulties with accessing the complex, fluid-388 filled cavity of the inner ear within the dense temporal bone, currently hampering diagnosis and 389 targeted drug delivery, may be behind us with the advent of microneedle technologies for safe 390 sampling of inner ear fluids and local treatments, as microlitres of perilymph fluid can be 391 extracted this way (68). The class of molecules that do not encode for proteins but regulate gene 392 expression, non-coding RNA, are revealing an increasing role in cellular process, including 393 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 18 differentiation and maturation, and their dysregulation can cause cancer. Perhaps renewed 394 interest in solving the OTSC loci should include looking for non-coding RNA targets. 395

396 Ethics Statement 397 Prior approval to study live research participants with HL and their blood relatives was granted 398 by the regional ethics authority (Hearing Loss Project #01.186, Human Research Ethics Board, 399 St John’s, NL, Canada). Informed consent from research participants was by written consent. 400 Clinical recruitment, pedigree structure and classification of HL 401 Research participants underwent audiological and medical examinations, permitted access to 402 their medical records, and completed a medical questionnaire (19). HL was classified based on 403 the pure-tone threshold averages of 0.5, 1.0 and 2.0 kHz, as defined by the American Speech and 404 Hearing Association (AHSA). Family members were assessed by our clinical team to update 405 audiograms and confirm middle ear status. A difference of >10 dB HL between air and bone 406 conduction sensitivity represented a significant conductive component associated with impaired 407 sound transfer through the middle ear. Conservative clinical criteria were used to assign 408 otosclerosis affection status: affected were blood relatives with surgically confirmed otosclerosis 409 at any age; unaffected were blood relatives ≥ 55 years of age with normal bilateral hearing 410 thresholds. Of nine siblings in generation III, five have otosclerosis (Fig 2). The father (PID II-7) 411 reportedly had normal hearing but was "very sensitive to noise" and the mother (PID II-8) had 412 HL (uncategorized) along with two of her nine siblings (PID II-16, PID II-17) who are reported 413 to have age-related HL. X-linked inheritance could be ruled out as otosclerosis is not more 414 severe in males; however, without further clinical information, the inheritance pattern is 415 consistent with both AD and AR inheritance. 416 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 19 Test for linkage to mapped OTSC loci-haplotype sharing and targeted sequencing 417 Genomic DNA was extracted from peripheral blood (69). Informative microsatellite markers (4-418 10 per locus) spanning seven OTSC loci and two otosclerosis-suceptibility genes including 419 OTSC1 (15q), OTSC2 (7q), OTSC3 (6p), OTSC4 (16q), OTSC5 (3q), OTSC7 (6q), OTSC8 (9p) 420 and genes COL1A1 (17q) and NOG (17q) were genotyped and shared alleles noted among 421 affected siblings. Primers were labelled (6-FAM), amplified using touchdown PCR, size 422 fractionated (ABI PRISM model 3130xl) and analyzed with Gene Mapper software (v4.0). 423 Haplotypes were recapitulated according to “least recombination rules” and paternal haplotypes 424 were inferred due to absence of DNA. Full gene sequencing (Sanger) was performed on three 425 affected (PIDs III-1, III-3, III-9) and one unaffected (PID III-7) for select annotated genes 426 [March 2006 assembly (NCBI build 36.1)] within OTSC loci with disease-associated haplotypes. 427 Genes were selected for sequencing if they had a functional role in bone remodeling or immune 428 response. Primers were designed to amplify the longest isoform, including all intron/exon 429 boundaries and UTRs and purified PCR products were bidirectionally sequenced (Big Dye 430 Terminator V3.1 kit; ABI PRISM 3130xl DNA Analyzer). We used Mutation Surveyor software 431 (version 4.07, SoftGenetics LLC State College, PA 16803) to select quality reads and analyze 432 DNA sequences. Rare sequencing variants (<2 %) were subjected to in silico tools (SIFT, 433 PolyPhen, Human Splicing Finder (HSF), MaxEntScan, NNSLICE, GeneSplicer, Known 434 constitutive signals) to predict pathogenicity. Variants were filtered out if they were absent in 435 affected or present in unaffected relatives. The proband was also screened for the 15-bp coding 436 deletion in FOXL1 (rs764026385; OTSC11) identified in an NL family (70) and for the rare 437 otosclerosis variants in SERPINF1 (71). All remaining variants were subjected to cascade 438 sequencing to verify co-segregation with otosclerosis in the family. 439 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 20 Genome wide SNP genotyping and multipoint linkage analysis 440 As a targeted genetic analysis did not identify the otosclerosis gene and the inheritance pattern is 441 not clear, genome wide SNP genotyping and multipoint linkage analysis were performed under 442 both AR and AD models with Merlin (version 1.1.2) (72, 73), assuming complete (100%) 443 penetrance. LOD scores were calculated at recombination fractions of 0.000 to 0.5000. We used 444 the 610K Illumina SNP array (Genome Centre, McGill University, QC, Canada) on samples 445 from five affected siblings (PIDs III-1, III-3, III-6, III-8, III-9), an unaffected sibling (PID III-7), 446 a maternal sibling with HL (PID II-17) and a maternal sibling with normal hearing (PID II-15). 447 As well, nine population control samples with normal hearing were used to estimate minor allele 448 frequencies of unavailable family members. Genotypes were analyzed at The Centre for Applied 449 Genomics (TCAG, University of Toronto, ON, Canada) and exported from GenomeStudio 450 software (v2010.3). 451 Sequencing, variant filtering and cascade sequencing under linked regions 452 Exome sequencing was carried out on four affected siblings (PIDs III-1, III-3, III-6, and III-9) 453 and two older controls (55, 60 yrs old) with normal hearing thresholds. Library preparation was 454 done with TrueSeq Prep Kit and samples run on the Illumina Hiseq 2000, generating 50-150 455 million 100-bp paired end reads. Reads >32-bp long were aligned to the 1000 genome reference 456 using Burrows-Wheeler Aligner (BWA) and merged with Picard software (Broad Institute). 457 Where multiple base mismatches and false positive variant calls were recorded, insertions and 458 deletions were realigned using GATK software (73, 74). The percentage of aligned region 459 coverage was detected using the Genome Centre's in-house database. The regions were identified 460 as high coverage (>400X), low coverage (<50X), low mean mapq MQ (<20X) and no data. Rare 461 variants were considered at a higher frequency (MAF<2%) to account for potential founder 462 effects. Rare variants, absent in one or more affected, present in one or more controls or also 463 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 21 identified in solved otosclerosis patients (e.g. FOXL1) were filtered out. Variants with a 464 minimum of 20X coverage were analysed with in silico tools including samtools mpileup 465 algorithm (75), SnpSift (76), SnpEff (76), SIFT, PolyPhen-2, PANTHER and ClustalW. In 466 addition, the functional consequence of KPNA7, c.49 C>T, p. R17X was analysed using Alamut 467 Visual Plus by Sophia Genetics, version 1.13 (2024). At this stage, allele frequencies of genetic 468 variants were checked in two research cohorts, the NL Osteoarthritis Study (NFOAS) consisting 469 of 1,000 total joint (knee and/or hip) replacement patients and the NL Colorectal Cancer 470 Registry, where samples were genotyped by Illumina microarray platforms and then imputed 471 with 1000 genome project data as reference panels. All remaining variants underwent cascade 472 screening and were tested against 38 otosclerosis probands (from ON Canada; Western 473 University ethics #103,679), 149 uncategorized HL probands (from NL) and 727 population 474 controls. 475 Web Resources 476 477 ASHA (Amer Speech-Language Hearing Assoc), https://www.asha.org/practice-portal/clinical-478 topics/hearing-loss/ 479 Burrows-Wheeler Aligner BWA, http://bio-bwa.sourceforge.net/ 480 ClinVar, https://www.ncbi.nlm.nih.gov/clinvar/ 481 ClustalW, https://www.genome.jp/tools-bin/clustalw 482 dbSNP, https://www.ncbi.nlm.nih.gov/snp/ 483 gnomAD, https://gnomad.broadinstitute.org/ 484 Hereditary Hearing Loss homepage, http://hereditaryhearingloss.org/ 485 Human Splicing Finder, www.umd.be/HSF3/ 486 487 1000 genomes, https://www.internationalgenome.org/home 488 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 22 Online Mendelian Inheritance in Man, http://omim.org 489 Primer3, https://bioinfo.ut.ee/primer3-0.4.0/ 490 PANTHER, http://www.pantherdb.org/ 491 PolyPhen-2, http://genetics.bwh.harvard.edu/pph2/ 492 RefSeq, https://www.ncbi.nlm.nih.gov/refseq/ 493 SIFT, https://sift.bii.a-star.edu.sg/ 494 SNP database, http://www.ncbi.nlm.nih.gov/projects/SNP/ 495 The MathWorks, Inc. https://www.mathworks.com 496 Varsome, http://varsome.com 497 UCSC Genome Browser, https://genome.ucsc.edu/ 498

499 This study was funded by the Canadian Institutes of Health Research (#222294), Canadian 500 Foundation for Innovation (#9384, #13120), and Genome Canada/Genome Atlantic (AMGGI) to 501 T.L.Y. Support was also provided by Memorial University, Town of Grand Falls-Windsor 502 (Excite Corporation) and the Government of Newfoundland and Labrador. One of the first 503 authors, N.A., is a recipient of a CIHR Fellowship and this paper includes contributions from 504 their PhD research. 505 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 23

506 507 1. Sönmez S, Orhan KS, Baumgartner W-D, Zarowski A, Özgirgin NO, Barbara M, et al., 508 editors. the Fifth International Symposium on Otosclerosis and Stapes Surgery. Symposium on 509 Otosclerosis and Stapes Surgery J Int Adv Otol; 2025. 510 2. Capobianco S, Lazzerini F, Bruschini L, Fiacchini G, Forli F. Exploring the genetic 511 landscape of otosclerosis: current understanding and future perspectives. Acta Otorhinolaryngol 512 Ital. 2025;45(Suppl. 1):S2-s17. 513 3. Mudry A. Adam Politzer (1835-1920) and the description of otosclerosis. Otol Neurotol. 514 2006;27(2):276-81. 515 4. Rämö JT, Kiiskinen T, Seist R, Krebs K, Kanai M, Karjalainen J, et al. Genome-wide 516 screen of otosclerosis in population biobanks: 27 loci and shared associations with skeletal 517 structure. Nat Commun. 2023;14(1):157. 518 5. Huybrechts Y, Mortier G, Boudin E, Van Hul W. WNT Signaling and Bone: Lessons 519 From Skeletal Dysplasias and Disorders. Front Endocrinol (Lausanne). 2020;11:165. 520 6. Sanverdi SE, Ozgen B, Dolgun A, Sarac S. Incomplete endochondral ossification of the 521 otic capsule, a variation in children: evaluation of its prevalence and extent in children with and 522 without sensorineural hearing loss. AJNR Am J Neuroradiol. 2015;36(1):171-5. 523 7. Frisch T, Sørensen MS, Overgaard S, Lind M, Bretlau P. Volume-Referent Bone 524 Turnover Estimated From the Interlabel Area Fraction After Sequential Labeling. Bone. 525 1998;22(6):677-82. 526 8. Moser T, Veillon F, Sick H, Riehm S. The hypodense focus in the petrous apex: a 527 potential pitfall on multidetector CT imaging of the temporal bone. AJNR Am J Neuroradiol. 528 2008;29(1):35-9. 529 9. Hawke M, Jahn AF. Bone formation in the normal human otic capsule. Arch 530 Otolaryngol. 1975;101(8):462-4. 531 10. Richard C, Doherty JK, Fayad JN, Cordero A, Linthicum FH, Jr. Identification of target 532 proteins involved in cochlear otosclerosis. Otol Neurotol. 2015;36(5):923-31. 533 11. Wang PC, Merchant SN, McKenna MJ, Glynn RJ, Nadol JB, Jr. Does otosclerosis occur 534 only in the temporal bone? Am J Otol. 1999;20(2):162-5. 535 12. Zehnder AF, Kristiansen AG, Adams JC, Merchant SN, McKenna MJ. Osteoprotegerin 536 in the inner ear may inhibit bone remodeling in the otic capsule. Laryngoscope. 537 2005;115(1):172-7. 538 13. Hansen LJ, Bloch SL, Sørensen MS. Cellular voids in the pathogenesis of otosclerosis. 539 Acta Otolaryngol. 2023;143(3):250-3. 540 14. Jáuregui EJ, Akil O, Acevedo C, Hall-Glenn F, Tsai BS, Bale HA, et al. Parallel 541 mechanisms suppress cochlear bone remodeling to protect hearing. Bone. 2016;89:7-15. 542 15. Mallik S, Poch D, Burick S, Schlieker C. Protein folding and quality control during 543 nuclear transport. Curr Opin Cell Biol. 2024;90:102407. 544 16. Paciorkowski AR, Weisenberg J, Kelley JB, Spencer A, Tuttle E, Ghoneim D, et al. 545 Autosomal recessive mutations in nuclear transport factor KPNA7 are associated with infantile 546 spasms and cerebellar malformation. Eur J Hum Genet. 2014;22(5):587-93. 547 17. Wang W, Miyamoto Y, Chen B, Shi J, Diao F, Zheng W, et al. Karyopherin α deficiency 548 contributes to human preimplantation embryo arrest. J Clin Invest. 2023;133(2). 549 18. Martin TJ. Parathyroid Hormone-Related Protein, Its Regulation of Cartilage and Bone 550 Development, and Role in Treating Bone Diseases. Physiol Rev. 2016;96(3):831-71. 551 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 24 19. Abdelfatah N. The genetic aetiology of otosclerosis in the population of Newfoundland 552 and Labrador: Memorial University of Newfoundland; 2014. 553 20. Oza AM, DiStefano MT, Hemphill SE, Cushman BJ, Grant AR, Siegert RK, et al. Expert 554 specification of the ACMG/AMP variant interpretation guidelines for genetic hearing loss. Hum 555 Mutat. 2018;39(11):1593-613. 556 21. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, et al. Standards and 557 guidelines for the interpretation of sequence variants: a joint consensus recommendation of the 558 American College of Medical Genetics and Genomics and the Association for Molecular 559 Pathology. Genet Med. 2015;17(5):405-24. 560 22. Van Den Bogaert K, Govaerts PJ, Schatteman I, Brown MR, Caethoven G, Offeciers FE, 561 et al. A second gene for otosclerosis, OTSC2, maps to chromosome 7q34-36. Am J Hum Genet. 562 2001;68(2):495-500. 563 23. Schrauwen I, Ealy M, Huentelman MJ, Thys M, Homer N, Vanderstraeten K, et al. A 564 genome-wide analysis identifies genetic variants in the RELN gene associated with otosclerosis. 565 Am J Hum Genet. 2009;84(3):328-38. 566 24. Vuorinen E. Nuclear import protein KPNA7 and its cargos: Diverse roles in the 567 regulation of cancer cell growth, mitosis and nuclear morphology. 2018. 568 25. Bloch SL. On the biology of the bony otic capsule and the pathogenesis of otosclerosis. 569 Dan Med J. 2012;59(10):B4524. 570 26. Bloch SL, Sørensen MS. The role of connectivity and stochastic osteocyte behavior in 571 the distribution of perilabyrinthine bone degeneration. A Monte Carlo based simulation study. 572 Hear Res. 2016;335:1-8. 573 27. Frisch T, Bloch SL, Sørensen MS. Prevalence, size and distribution of microdamage in 574 the human otic capsule. Acta Oto-Laryngologica. 2015;135(8):771-5. 575 28. Hansen LJ, Bloch SL, Sørensen MS. Identification of cellular voids in the human otic 576 capsule. Journal of the Association for Research in Otolaryngology. 2021;22(5):591-9. 577 29. Hansen LJ, Bloch SL, Frisch T, Sørensen MS. Microcrack surface density in the human 578 otic capsule: An unbiased stereological quantification. The Anatomical Record. 579 2021;304(5):961-7. 580 30. Declau F, van den Bogaert K, Van de Heyning P, Offeciers E, Govaerts P, van Camp G. 581 Phenotype-genotype correlations in otosclerosis: clinical features of OTSC2. Adv 582 Otorhinolaryngol. 2007;65:114-8. 583 31. Lange A, Mills RE, Lange CJ, Stewart M, Devine SE, Corbett AH. Classical nuclear 584 localization signals: definition, function, and interaction with importin α. Journal of Biological 585 Chemistry. 2007;282(8):5101-5. 586 32. Kelley JB, Talley AM, Spencer A, Gioeli D, Paschal BM. Karyopherin alpha7 (KPNA7), 587 a divergent member of the importin alpha family of nuclear import receptors. BMC Cell Biol. 588 2010;11:63. 589 33. Kobe B. Autoinhibition by an internal nuclear localization signal revealed by the crystal 590 structure of mammalian importin alpha. Nat Struct Biol. 1999;6(4):388-97. 591 34. Sekimoto T, Yoneda Y. Intrinsic and extrinsic negative regulators of nuclear protein 592 transport processes. Genes to Cells. 2012;17(7):525-35. 593 35. Lam MH, Thomas RJ, Martin TJ, Gillespie MT, Jans DA. Nuclear and nucleolar 594 localization of parathyroid hormone-related protein. Immunol Cell Biol. 2000;78(4):395-402. 595 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 25 36. Grayeli AB, Sterkers O, Roulleau P, Elbaz P, Ferrary E, Silve C. Parathyroid hormone-596 parathyroid hormone-related peptide receptor expression and function in otosclerosis. Am J 597 Physiol. 1999;277(6):E1005-12. 598 37. Pumroy RA, Cingolani G. Diversification of importin-α isoforms in cellular trafficking 599 and disease states. Biochem J. 2015;466(1):13-28. 600 38. Oostdyk LT, McConnell MJ, Paschal BM. Characterization of the Importin-β binding 601 domain in nuclear import receptor KPNA7. Biochem J. 2019;476(21):3413-34. 602 39. Sharif M, Detti L, Van den Veyver IB. Take your mother's ferry: preimplantation embryo 603 development requires maternal karyopherins for nuclear transport. J Clin Invest. 2023;133(2). 604 40. Tejomurtula J, Lee K-B, Tripurani SK, Smith GW, Yao J. Role of importin alpha8, a new 605 member of the importin alpha family of nuclear transport proteins, in early embryonic 606 development in cattle. Biology of reproduction. 2009;81(2):333-42. 607 41. Liu X, Chen J, Long X, Lan J, Liu X, Zhou M, et al. RSL1D1 promotes the progression 608 of colorectal cancer through RAN-mediated autophagy suppression. Cell Death & Disease. 609 2022;13(1):43. 610 42. Miyamoto Y, Yamada K, Yoneda Y. Importin α: a key molecule in nuclear transport and 611 non-transport functions. J Biochem. 2016;160(2):69-75. 612 43. Cingolani G, Bednenko J, Gillespie MT, Gerace L. Molecular basis for the recognition of 613 a nonclassical nuclear localization signal by importin beta. Mol Cell. 2002;10(6):1345-53. 614 44. Chen H, Tan XN, Hu S, Liu RQ, Peng LH, Li YM, et al. Molecular Mechanisms of 615 Chondrocyte Proliferation and Differentiation. Front Cell Dev Biol. 2021;9:664168. 616 45. Karaplis AC, Luz A, Glowacki J, Bronson RT, Tybulewicz V, Kronenberg HM, et al. 617 Lethal skeletal dysplasia from targeted disruption of the parathyroid hormone-related peptide 618 gene. Genes & development. 1994;8(3):277-89. 619 46. Tanaka N, Ohno S, Honda K, Tanimoto K, Doi T, Ohno-Nakahara M, et al. Cyclic 620 mechanical strain regulates the PTHrP expression in cultured chondrocytes via activation of the 621 Ca2+ channel. Journal of dental research. 2005;84(1):64-8. 622 47. Wu Q, Zhang Y, Chen Q. Indian hedgehog is an essential component of 623 mechanotransduction complex to stimulate chondrocyte proliferation. J Biol Chem. 624 2001;276(38):35290-6. 625 48. Chen Y, Mehmood K, Chang Y-F, Tang Z, Li Y, Zhang H. The molecular mechanisms 626 of glycosaminoglycan biosynthesis regulating chondrogenesis and endochondral ossification. 627 Life Sciences. 2023;335:122243. 628 49. Hallett SA, Matsushita Y, Ono W, Sakagami N, Mizuhashi K, Tokavanich N, et al. 629 Chondrocytes in the resting zone of the growth plate are maintained in a Wnt-inhibitory 630 environment. Elife. 2021;10:e64513. 631 50. Tong W, Xu J, Qi Q, Chen H, Huang T, Chen C, et al. PTHrP buffers Wnt/β-catenin 632 activity through a negative feedback loop to maintain articular cartilage homeostasis. bioRxiv. 633 2022:2022.11. 25.517940. 634 51. Schipani E, Lanske B, Hunzelman J, Luz A, Kovacs CS, Lee K, et al. Targeted 635 expression of constitutively active receptors for parathyroid hormone and parathyroid hormone-636 related peptide delays endochondral bone formation and rescues mice that lack parathyroid 637 hormone-related peptide. Proc Natl Acad Sci U S A. 1997;94(25):13689-94. 638 52. Al-Jallad H, Palomo T, Roughley P, Glorieux FH, McKee MD, Moffatt P, et al. The 639 effect of SERPINF1 in-frame mutations in osteogenesis imperfecta type VI. Bone. 2015;76:115-640 20. 641 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 26 53. Schrauwen I, Valgaeren H, Tomas-Roca L, Sommen M, Altunoglu U, Wesdorp M, et al. 642 Variants affecting diverse domains of MEPE are associated with two distinct bone disorders, a 643 craniofacial bone defect and otosclerosis. Genet Med. 2019;21(5):1199-208. 644 54. Hawkey-Noble A, Pater JA, Kollipara R, Fitzgerald M, Maekawa AS, Kovacs CS, et al. 645 Mutation of foxl1 Results in Reduced Cartilage Markers in a Zebrafish Model of Otosclerosis. 646 Genes (Basel). 2022;13(7). 647 55. Drabkin M, Jean MM, Noy Y, Halperin D, Yogev Y, Wormser O, et al. SMARCA4 648 mutation causes human otosclerosis and a similar phenotype in mice. J Med Genet. 649 2024;61(2):117-24. 650 56. Fabbris C, Molteni G, Tommasi N, Marchioni D. Does pregnancy have an influence on 651 otosclerosis? J Laryngol Otol. 2022;136(3):191-6. 652 57. Kovacs CS. Calcium and bone metabolism in pregnancy and lactation. J Clin Endocrinol 653 Metab. 2001;86(6):2344-8. 654 58. Kovacs CS. Calcium and bone metabolism during pregnancy and lactation. J Mammary 655 Gland Biol Neoplasia. 2005;10(2):105-18. 656 59. Kovacs CS. Maternal Mineral and Bone Metabolism During Pregnancy, Lactation, and 657 Post-Weaning Recovery. Physiol Rev. 2016;96(2):449-547. 658 60. Mardinian K, Adashek JJ, Botta GP, Kato S, Kurzrock R. SMARCA4: implications of an 659 altered chromatin-remodeling gene for cancer development and therapy. Molecular cancer 660 therapeutics. 2021;20(12):2341-51. 661 61. Schoenfeld AJ, Bandlamudi C, Lavery JA, Montecalvo J, Namakydoust A, Rizvi H, et al. 662 The genomic landscape of SMARCA4 alterations and associations with outcomes in patients 663 with lung cancer. Clinical Cancer Research. 2020;26(21):5701-8. 664 62. Tian Y, Xu L, Li X, Li H, Zhao M. SMARCA4: Current status and future perspectives in 665 non-small-cell lung cancer. Cancer letters. 2023;554:216022. 666 63. Chen A, Zhong L, Lv J. FOXL1 overexpression is associated with poor outcome in 667 patients with glioma. Oncology Letters. 2019;18(1):751-7. 668 64. Vuorinen EM, Rajala NK, Ihalainen TO, Kallioniemi A. Depletion of nuclear import 669 protein karyopherin alpha 7 (KPNA7) induces mitotic defects and deformation of nuclei in 670 cancer cells. BMC cancer. 2018;18(1):325. 671 65. Laurila E, Vuorinen E, Savinainen K, Rauhala H, Kallioniemi A. KPNA7, a nuclear 672 transport receptor, promotes malignant properties of pancreatic cancer cells in vitro. Exp Cell 673 Res. 2014;322(1):159-67. 674 66. Diaz C, Thankam FG, Agrawal DK. Karyopherins in the Remodeling of Extracellular 675 Matrix: Implications in Tendon Injury. J Orthop Sports Med. 2023;5(3):357-74. 676 67. Mauri G, Sartore‐Bianchi A, Russo AG, Marsoni S, Bardelli A, Siena S. Early‐onset 677 colorectal cancer in young individuals. Molecular oncology. 2019;13(2):109-31. 678 68. Tavazzani E, Spaiardi P, Contini D, Sancini G, Russo G, Masetto S. Precision medicine: 679 a new era for inner ear diseases. Front Pharmacol. 2024;15:1328460. 680 69. Miller SA, Dykes DD, Polesky HF. A simple salting out procedure for extracting DNA 681 from human nucleated cells. Nucleic Acids Res. 1988;16(3):1215. 682 70. Abdelfatah N, Mostafa AA, French CR, Doucette LP, Penney C, Lucas MB, et al. A 683 pathogenic deletion in Forkhead Box L1 (FOXL1) identifies the first otosclerosis (OTSC) gene. 684 Human genetics. 2022;141(3):965-79. 685 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 27 71. Ziff JL, Crompton M, Powell HR, Lavy JA, Aldren CP, Steel KP, et al. Mutations and 686 altered expression of SERPINF1 in patients with familial otosclerosis. Hum Mol Genet. 687 2016;25(12):2393-403. 688 72. Abecasis GR, Cherny SS, Cookson WO, Cardon LR. Merlin--rapid analysis of dense 689 genetic maps using sparse gene flow trees. Nat Genet. 2002;30(1):97-101. 690 73. DePristo MA, Banks E, Poplin R, Garimella KV, Maguire JR, Hartl C, et al. A 691 framework for variation discovery and genotyping using next-generation DNA sequencing data. 692 Nat Genet. 2011;43(5):491-8. 693 74. Van der Auwera GA, Carneiro MO, Hartl C, Poplin R, Del Angel G, Levy-Moonshine A, 694 et al. From FastQ data to high confidence variant calls: the Genome Analysis Toolkit best 695 practices pipeline. Curr Protoc Bioinformatics. 2013;43(1110):11.0.1-.0.33. 696 75. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, et al. The Sequence 697 Alignment/Map format and SAMtools. Bioinformatics. 2009;25(16):2078-9. 698 76. Cingolani P, Platts A, Wang le L, Coon M, Nguyen T, Wang L, et al. A program for 699 annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the 700 genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin). 2012;6(2):80-92. 701 702 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 28 703 704 Fig 1. Pre- and post-stapedectomy audiograms reveal successful surgical treatment of 705 bilateral otosclerosis in the proband (PID III-1). 706 Pre-surgery audiogram revealed a conductive loss with borderline cochlear (sensorineural) HL. 707 Post stapedectomy of the right ear improved hearing and the air-bone gap was mainly resolved 708 but some HL remained, especially in the low frequencies. After stapedectomy of the left ear at 709 age 44, hearing improved across all frequencies with only a mild sensorineural HL remaining in 710 the mid to high frequencies. At age 51, the proband experienced mild hearing loss in the low and 711 mid frequencies of the left ear but high frequencies showed moderate to severe loss at 4000 Hz 712 and 8000 Hz, likely due to disease processes on the cochlear side of the round window. Pre = 713 pre-stapedectomy audiogram, age 42; post RT = post-stapedectomy audiogram, right ear, age 42; 714 post bil 1 = first post-bilateral stapedectomy audiogram, age 44; post bil 2 = second post-715 bilateral stapedectomy audiogram, age 51;  = unmasked bone conduction; □ = masked bone 716 conduction. Generated by MathWorks. (2020). MATLAB (Version R2020a)[Computer software]. 717 718 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 29 719 720 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 30 Fig 2. Pedigree and otosclerosis-associated haplotype of a multiplex family from NL with 721 clinically confirmed otosclerosis. 722 (A) Pedigree of NL family with AD otosclerosis co-segregating with the KPNA7, c.49C>T 723 premature stop mutation and transmitted from the paternal side (orange haplotype). KPNA7 724 maps to the OTSC2 locus in the vicinity of RELN. (B) Sequencing electropherogram generated 725 from Mutation Surveyor software showing the heterozygous nonsense mutation. We used 726 diamonds instead of squares (males) and circles (females) to protect the identity of research 727 participants. 728 729 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint Manuscript: KPNA7 causes otosclerosis Benteau et al., 2025 31 730 Fig 3. New mechanistic model for otosclerosis due to pathogenic mutations in KPNA7 and 731 other OTSC genes within the aging otic capsule. 732 Otosclerosis genes (*) and otosclerosis susceptibility genes (green circles) are involved in 733 endochondral ossification signaling pathways. We hypothesize that KPNA7 haploinsufficiency 734 allows the nuclear import of PTHrP into the quiescent cells of globuli interossei trapped within 735 cellular voids of the aging otic capsule, re-activating endochondral ossification pathways in these 736 embryonic tissues. Created in BioRender. BENTEAU, T. 737 (2025) https://BioRender.com/mfhqhzv. 738 .CC-BY 4.0 International licenseavailable under a (which was not certified by peer review) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity. It is made The copyright holder for this preprintthis version posted August 26, 2025. ; https://doi.org/10.1101/2025.08.22.671726doi: bioRxiv preprint