Mutations in COCH that result in non-syndromic autosomal dominant deafness (DFNA9) affect matrix deposition of cochlin
- First Online:
- Cite this article as:
- Grabski, R., Szul, T., Sasaki, T. et al. Hum Genet (2003) 113: 406. doi:10.1007/s00439-003-0992-7
- 133 Views
The COCH gene mutated in autosomal dominant sensorineural deafness (DFNA9) encodes cochlin, a major constituent of the inner ear extracellular matrix. Sequence analysis of cochlin from DFNA9 patients identified five distinct single-amino-acid mutations within a conserved region (the LCCL domain) of cochlin. To define the molecular basis of DFNA9, we have generated myc-tagged wild-type and mutant cochlins and explored their behavior in transient transfection systems. Western blotting of cell lysates and culture media indicates that wild-type and mutant cochlins are synthesized and secreted in similar amounts. Immunofluorescent staining confirms that all are detected within the endoplasmic reticulum and the Golgi complex of transfected cells. Our findings suggest that COCH mutations are unlikely to cause abnormalities in secretion and suggest that extracellular events might cause DFNA9 pathology. In agreement, we show that wild-type cochlin accumulates in extracellular deposits that closely parallel the matrix component fibronectin, whereas mutant cochlins vary in the amount and pattern of extracellular material. Whereas some mutants exhibit an almost normal deposition pattern, some show complete lack of deposition. Our results suggest that DFNA9 results from gene products that fail to integrate correctly into the extracellular matrix. The partial or complete penetrance of integration defects suggests that DFNA9 pathology may be caused by multiple molecular mechanisms, including compromised ability of cochlin to self-assemble or to form appropriate complexes with other matrix components.
Hearing impairment is the most common human communication disorder, and a familial component is known to be involved in its etiology. Rapid advances in molecular and genetic technologies have led to the identification of an ever-growing number of genes whose products are essential for hearing. Mutations in the COCH (coagulation factor C homology) gene have been shown to correlate with DFNA9, a late-onset autosomal dominant non-syndromic hearing disorder (Verhagen et al. 2000, 2001; Kamarinos et al. 2001; Verstreken et al. 2001; Fransen et al. 1999; de Kok et al. 1999; Robertson et al. 1998; Manolis et al. 1996). Hearing loss begins in the 4th or 5th decade of life and initially involves the high frequencies (Versreken et al. 2001). Deafness is progressive and is usually complete by the 6th decade. In addition to cochlear involvement, DFNA9 patients also exhibit a spectrum of clinical vestibular dysfunctions. Penetrance of the vestibular symptoms is often incomplete, and some patients are minimally affected, whereas others suffer from repeated episodes of vertigo and severe balance disturbances (Verhagen et al. 2001).
Analysis of the COCH gene in families afflicted with DFNA9 has identified five different mutations in the LCCL domain, the only region showing alterations (de Kok et al. 1999; Fransen et al. 1999; Kamarinos et al. 2001; Verhagen et al. 2001; Verstreken et al. 2001; Robertson et al. 1998). All are single-base-pair substitutions that result in an amino acid change in residues that are identical in human, cow, mouse, and chicken (except for a conserved substitution of valine 66 to isoleucine in the chicken sequence). In vivo, the mutations influence only the larger isoform of cochlin, since the smaller isoforms do not contain the LCCL domain (Ikezono et al. 2001). Structural analysis of bacterially expressed LCCL fragment of wild-type and mutant cochlins indicates that four of the five described mutations (excluding the W117R mutation) disrupt the normal structure and lead to misfolding of the domain (Liepinsh et al. 2001). The consequences of such misfolding and the way in which they result in DFNA9 pathology have not been explored.
Here, we have explored whether mutations in cochlin affect its synthesis, secretion, and post-secretion behavior. Our data suggest a DFNA9 mechanism in which mutant cochlins do not correctly integrate into the ECM, leading to structural changes in the supportive matrix of the cochlear and vestibular systems. Such ECM changes are likely to have pronounced effects on the sensory function of both organs.
Materials and methods
Reagents and antibodies
Restriction enzymes and molecular reagents were from Promega (Madison, Wis.), New England BioLabs (Beverly, Mass.), or Qiagen (Chatsworth, Calif.). SuperSignal West Pico Chemiluminescence Substrate was from Pierce (Rockford, Ill.). Rabbit polyclonal antibodies against p115 and mouse polyclonal antibodies against GM130 have been described previously (Barroso et al. 1995; Nelson et al. 1998). Rabbit polyclonal anti-myc antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, Calif.), and monoclonal anti-myc antibodies were obtained from Invitrogen (Carlsbad, Calif.). Goat anti-rat and anti-mouse antibodies conjugated with Oregon Green or Texas Red-X and Alexa Fluor 594 Phalloidin were purchased from Molecular Probes (Eugene, Ore.). Horseradish peroxidase (HRP)-labeled sheep anti-rabbit IgG antibody, HRP-labeled donkey anti-mouse IgG antibody, and HRP-labeled rabbit anti-goat IgG antibody were from Amersham Life Science (Buckinghamshire, England). Antibodies against calreticulin were from Affinity BioReagents (Golden, Colo.). Antibodies against fibronectin, actinin, paxilin, and vinculin were kindly provided by Dr. Anne Woods (UAB).
Gene mutation nomenclature used in this article follows the recommendations of den Dunnen and Antonarakis (2001). Gene symbols used in this article follow the recommendations of the HUGO Gene Nomenclature Committee (Povey et al. 2001).The authors have made every attempt to perform the study in accordance with the recommendations made by Cooper et al. (2002).
Generation of constructs
Full-length human cochlin cDNA was provided by Dr. Rupert Timpl and subcloned first into pcDNA3.1(+) and then into pcDNA4/TO/Myc-His, both from Invitrogen (Carlsbad, Calif.). Point mutations were introduced with Quick Change Site-Directed Mutagenesis Kit, according to the manufacturer's protocol (Stratagene, La Jolla, Calif.). All mutated cochlin sequences were verified by sequencing.
In vitro transcription/translation
Cochlin was generated in vitro by using the TNT T7 Coupled Reticulocyte Lysate System from Promega, according to the manufacturer's directions. An aliquot of 2 μg cochlin-pBS cDNA was used in a total reaction volume of 100 μl. Trans 35S Label from ICN Pharmaceuticals was used as the source of 35S-methionine and 35S-cysteine. The reaction was carried out at 30°C for 4 h.
HeLa cells were grown in Dulbecco's modified Eagle's medium with glucose and glutamine (Mediatech, Comprehensive Cancer Center of the University of Alabama at Birmingham), supplemented with 10% fetal bovine serum (FBS, Life Technologies, Grand Island, N.Y.), 100 U/ml penicillin, and 100 μg/ml streptomycin (Life Technologies).
Transfections and immunofluorescence microscopy
Cells were transfected with the Calcium Phosphate Transfection System (Life Technologies) or with TransIT Polyamine Transfection Reagents (Mirus Corporation, Madison, Wis.), according to the manufacturer's protocols. Some 18–48 h after transfection, cells were fixed in 3% paraformaldehyde in phosphate-buffered saline (PBS) for 10 min at room temperature. Paraformaldehyde was quenched with 10 mM ammonium chloride, and cells were permeabilized with PBS, 0.1% Triton X-100, for 7 min at room temperature. In some experiments, cells were not permeabilized. The coverslips were washed (3×2 min) with PBS, then blocked in PBS, 0.4% fish skin gelatin, 0.2% Tween 20 for 5 min, followed by blocking in PBS, 2.5% goat serum, 0.2% Tween 20 for 5 min. Cells were incubated with primary antibody diluted in PBS, 0.4% fish skin gelatin, 0.2% Tween 20 for 45 min at 37°C. Coverslips were washed (5×5 min) with PBS, 0.2% Tween 20. Secondary antibodies coupled to Oregon Green or Texas red-X were diluted in 2.5% goat serum and incubated on coverslips for 30 min at 37°C. Coverslips were washed as above and mounted on slides in 9:1 glycerol:PBS with 0.1% q-phenylenediamine. Fluorescence patterns were visualized with a Leitz Orthoplan epifluorescence microscope (Wetzlar, Germany). Optical sections were captured with a charge-coupled-device high-resolution camera equipped with a camera/computer interface. Images were analyzed with a power Mac by using IPLab Spectrum software (Scanalytics, Fairfax, Va.). Actin filaments were visualized by labeling with Alexa Fluor 594 Phalloidin (Molecular Probes) diluted with 0.05% Tween 20 in PBS. Nuclei were counterstained with Hoechst 33258.
Immunoprecipitation, SDS-polyacrylamide gel electrophoresis, and immunoblotting
HeLa cells were solubilized in RIPA buffer (150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 50 mM TRIS, pH 8.0) supplemented with Complete Protease Inhibitors (Roche) 24 h after transfection. After incubation at 4°C for 30 min, samples were centrifuged at 16,000 g for 15 min. Supernatants were incubated at 4°C with 2 μg anti-myc antibodies for 2 h, followed by 20 μl 50% (v/v) protein A-Sepharose 4FF for 1 h. Beads were recovered by centrifugation and then washed four times with RIPA buffer containing inhibitors. Precipitates were analyzed by SDS-polyacrylamide gel electrophoresis (SDS-PAGE). Following SDS-PAGE, proteins were transferred to NitroPure nitrocellulose membrane (Micron Separations, Westborough, Mass.), and the membrane subjected to immunoblotting as previously described (Gao et al. 1998).
Protein N-glycosidase treatment
HeLa cells were lysed in RIPA buffer with protease inhibitors for 15 min on ice 24 h after transfection. Cell lysates were supplemented with 5 μl 10% NP-40, 5 μl 10× G7 buffer (NEB), and 1.5 μl protein N-glycosidase F (PNGase F), and incubated for 45 min at 37°C in 50 μl total volume. Samples were then supplemented with an equal volume of 2× Laemmli buffer, analyzed by SDS-PAGE, and immunoblotted with anti-myc antibodies.
Generation of cochlin constructs analogous to mutants that cause DFNA9
A full length clone encoding wild-type human cochlin was used as a template in polymerase chain reactions (PCRs) that introduced single-amino-acid substitutions into the coding region. The wild-type and mutant cochlins were then subcloned into a mammalian expression vector that added a myc/his-tag to the C-terminus of each protein (Fig. 1A).
Expression and secretion of wild-type and mutant cochlins
To determine whether mutant cochlins can be secreted from cells or are misfolded and removed through the endoplasmic-reticulum-associated degradation (ERAD) pathway, culture media were analyzed for cochlin. As shown in Fig. 2B, a band of ~69kDa is detected in culture media from cells expressing the wild-type or mutant cochlins, suggesting that the mutant proteins traverse the secretory pathway and are not extensively degraded. Only the I109 N mutant shows slightly reduced levels of secretion, possibly because of ERAD degradation or intracellular retention. The increase in size between the secreted and cellular cochlins (~69kDa versus ~63kDa) is attributable to processing of the N-linked oligosaccharide chains (data not shown).
Intracellular and extracellular localization of wild-type cochlin
Intracellular localization of mutant cochlins
Extracellular localization of mutant cochlins
The last decade has witnessed rapid progress in the discovery and characterization of genes involved in hereditary deafness (Petersen 2002; Steel and Bussoli 2001). COCH is responsible for post-lingual progressive autosomal-dominant hearing loss, DFNA9. Pathology of DFNA9 manifests as a sensorineural hearing loss starting in high frequencies and gradually leads to anacusis. DFNA9 is associated with significant histopathological changes in which acidophilic deposits are found in the peripheral auditory and vestibular systems. In patients with DFNA9 caused by the V66G, the G88E, and the W117R mutations, such accumulations of mucopolysaccharides have been found to obstruct the cochlear and vestibular nerve channels (Robertson et al. 1998; Khetarpal 2000; Merchant et al. 2000). Patients also show a significant decrease in the number of cells that normally express cochlin in the spiral lamina, spiral limbus, and spiral ligament, suggesting that the accumulation of deposits may be related to cell death.
The COCH gene product cochlin is a secretory protein that constitutes the major non-collagen component of the ECM of the inner ear (Ikezono et al. 2001). Cochlin contains two vWFA domains that are characteristic for proteins interacting with fibrillar collagens, suggesting a possible role in linking ECM components. Cochlin also contains a rare LCCL domain consisting of a central α-helix surrounded by six short β-sheets and stabilized by two di-sulfide bonds (Liepinsh et al. 2001; Fig. 1B). DFNA9 patients contain five distinct single-amino-acid substitutions, all located within the LCCL domain. All mutations, except the W117R substitution, result in misfolding of the LCCL domain. Interestingly, the W117R mutation causes the same histopathology as the "misfolded" V66G and G88E mutations, raising the possibility that different molecular mechanisms may lead to DFNA9 pathology.
To characterize the molecular basis of DFNA9 deafness, we generated mutant cochlins analogous to those found in DFNA9 patients and characterized their behavior in a transient transfection system.
Synthesis and secretion
We show that wild-type and mutant cochlins are synthesized as glycoproteins containing 2 N-linked carbohydrate chains in HeLa cells adapted to culture. Cochlin contains two asparagines that fit the NXS/T consensus sequence for glycosylation. Asp100 (NYS) lies within the LCCL domain, whereas Asp221 (NFT) is contained within the vWFA 1 region. Significantly, glycosylation of Asp100 appears unperturbed in all mutant cochlins, suggesting that the mutations do not fundamentally disrupt the global structure of the LCCL domain in vivo. Glycosylation of Asp221 is unlikely to be affected by mutations within the LCCL, since the LCCL and vWFA domains appear to fold autonomously (Liepinsh et al. 2001). That mutations within the LCCL domain do not cause global misfolding is supported by the finding that all mutant cochlins are secreted from cells. Misfolded secretory proteins are usually removed from the ER through retro-translocation and proteasome-mediated degradation (Hampton 2002). Some of the mutations may affect transit through the secretory pathway, since wild-type W117R and P51S show higher concentration within the Golgi than do the V66G, G88E, and I109 N mutants.
Our results show that DFNA9 is unlikely to be caused by significant defects in cochlin synthesis, glycosylation, or secretion and suggest that the acidophilic deposits and loss of cellularity may be linked to extracellular matrix-related events. In support, we document differences in the matrix association of wild-type and mutant cochlins.
The most drastic effects are observed with the V66G, G88E, and I109 N mutants. Despite being secreted from cells (Fig. 2B), these mutants proteins do not localize to the ECM; they might integrate into ECM in situ, by binding to fibrillar collagens or other ECM molecules through their vWFA domains. However, the use of HeLa cells that do not express collagen or other cochlea-specific matrix proteins has allowed the visualization of the deposition defect. The finding that LCCL domains containing the V66G or G88E mutations misfold and aggregate when expressed in bacteria (Liepinsh et al. 2001) has led to the suggestion that the accumulation of acidophilic deposits in DFNA9 patients might results from the intracellular or extracellular deposition of misfolded cochlin mutants. Our data do not support such a model: mutant cochlins were not found to aggregate intracellularly and were not detected in extracellular deposits. It is likely that the difference in misfolding lies in the specialized redox environment that is present within the mammalian ER and that supports di-sulfide bonds formation and protein folding (Fassio and Sitia 2002). Mutant cochlins may achieve a more native folding state in mammalian cells. Analogous conditions would be present in DFNA9 patients, suggesting that acute misfolding and aggregation of mutant cochlins are unlikely to account for DFNA9 histopathology.
This is also strongly suggested by the behavior of the W117R and P51S mutants. These mutant proteins form extracellular deposits that are indistinguishable from wild-type cochlin and yet cause the same histopathology as the V66G and G88E mutants. It is likely that, in situ, they are unable to interact with their cognate ECM components. Together, the data suggest that the LCCL domain may have multiple functions and that multiple protein-protein interactions may underlie cochlin localization and function in vivo. First, the LCCL domain appears to mediate cochlin deposition within the ECM, whereas the V66G, G88E, and I109 N mutations inhibit this process. Deposition is possibly linked to LCCL-mediated homo-oligomerization of cochlin or association with specific ECM molecules. Second, the LCCL domain is likely to mediate interactions with additional ECM components, with the W117R and P51S mutations inhibiting such events. These interactions may involve ECM molecules specific to the cochlea and the vestibular system. The data are consistent with cochlin acting as a master regulator that organizes and gate-keeps the specific architecture of ECM in the cochlear and vestibular systems. Alterations in cochlin ability to integrate into the ECM or to interact with specific ECM components may represent two molecular mechanisms that lead to DFNA9 deafness.
We thank Dr. Cynthia Morton and Nancy Robertson for helpful discussion and sharing of results prior to publication. We are grateful to Dr. Anne Woods for providing antibodies and explanations and to Dr. Ben Sha for modeling the LCCL domain.