Cadherin 23 and protocadherin 15 interact to form tip-link filaments in sensory hair cells.
ABSTRACT Hair cells of the inner ear are mechanosensors that transduce mechanical forces arising from sound waves and head movement into electrochemical signals to provide our sense of hearing and balance. Each hair cell contains at the apical surface a bundle of stereocilia. Mechanoelectrical transduction takes place close to the tips of stereocilia in proximity to extracellular tip-link filaments that connect the stereocilia and are thought to gate the mechanoelectrical transduction channel. Recent reports on the composition, properties and function of tip links are conflicting. Here we demonstrate that two cadherins that are linked to inherited forms of deafness in humans interact to form tip links. Immunohistochemical studies using rodent hair cells show that cadherin 23 (CDH23) and protocadherin 15 (PCDH15) localize to the upper and lower part of tip links, respectively. The amino termini of the two cadherins co-localize on tip-link filaments. Biochemical experiments show that CDH23 homodimers interact in trans with PCDH15 homodimers to form a filament with structural similarity to tip links. Ions that affect tip-link integrity and a mutation in PCDH15 that causes a recessive form of deafness disrupt interactions between CDH23 and PCDH15. Our studies define the molecular composition of tip links and provide a conceptual base for exploring the mechanisms of sensory impairment associated with mutations in CDH23 and PCDH15.
- SourceAvailable from: Thomas Effertz[Show abstract] [Hide abstract]
ABSTRACT: Identification of the auditory hair cell mechano-electrical transduction (hcMET) channel has been a major focus in the hearing research field since the 1980s when direct mechanical gating of a transduction channel was proposed (Corey and Hudspeth J Neurosci 3:962-976, 1983). To this day, the molecular identity of this channel remains controversial. However, many of the hcMET channel's properties have been characterized, including pore properties, calcium-dependent ion permeability, rectification, and single channel conductance. At this point, elucidating the molecular identity of the hcMET channel will provide new tools for understanding the mechanotransduction process. This review discusses the significance of identifying the hcMET channel, the difficulties associated with that task, as well as the establishment of clear criteria for this identification. Finally, we discuss potential candidate channels in light of these criteria.Pflügers Archiv - European Journal of Physiology 09/2014; · 3.07 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Background Usher syndrome is an autosomal recessive disease that associates sensorineural hearing loss, retinitis pigmentosa and, in some cases, vestibular dysfunction. It is clinically and genetically heterogeneous. To date, 10 genes have been associated with the disease, making its molecular diagnosis based on Sanger sequencing, expensive and time-consuming. Consequently, the aim of the present study was to develop a molecular diagnostics method for Usher syndrome, based on targeted next generation sequencing.MethodsA custom HaloPlex panel for Illumina platforms was designed to capture all exons of the 10 known causative Usher syndrome genes (MYO7A, USH1C, CDH23, PCDH15, USH1G, CIB2, USH2A, GPR98, DFNB31 and CLRN1), the two Usher syndrome-related genes (HARS and PDZD7) and the two candidate genes VEZT and MYO15A. A cohort of 44 patients suffering from Usher syndrome was selected for this study. This cohort was divided into two groups: a test group of 11 patients with known mutations and another group of 33 patients with unknown mutations.ResultsForty USH patients were successfully sequenced, 8 USH patients from the test group and 32 patients from the group composed of USH patients without genetic diagnosis. We were able to detect biallelic mutations in one USH gene in 22 out of 32 USH patients (68.75%) and to identify 79.7% of the expected mutated alleles. Fifty-three different mutations were detected. These mutations included 21 missense, 8 nonsense, 9 frameshifts, 9 intronic mutations and 6 large rearrangements.Conclusions Targeted next generation sequencing allowed us to detect both point mutations and large rearrangements in a single experiment, minimizing the economic cost of the study, increasing the detection ratio of the genetic cause of the disease and improving the genetic diagnosis of Usher syndrome patients.Orphanet Journal of Rare Diseases 11/2014; 9(1):168. · 3.96 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Loss of function variants in the PCDH15 gene can cause Usher syndrome type 1F, an autosomal recessive disease associated with profound congenital hearing loss, vestibular dysfunction, and retinitis pigmentosa. The Ashkenazi Jewish population has an increased incidence of Usher syndrome type 1F (founder variant p.Arg245X accounts for 75% of alleles), yet the variant spectrum in a panethnic population remains undetermined. We sequenced the coding region and intron-exon borders of PCDH15 using next-generation DNA sequencing technology in approximately 14,000 patients from fertility clinics. More than 600 unique PCDH15 variants (single nucleotide changes and small indels) were identified, including previously described pathogenic variants p.Arg3X, p.Arg245X (five patients), p.Arg643X, p.Arg929X, and p.Arg1106X. Novel truncating variants were also found, including one in the N-terminal extracellular domain (p.Leu877X), but all other novel truncating variants clustered in the exon 33 encoded C-terminal cytoplasmic domain (52 patients, 14 variants). One variant was observed predominantly in African Americans (carrier frequency of 2.3%). The high incidence of truncating exon 33 variants indicates that they are unlikely to cause Usher syndrome type 1F even though many remove a large portion of the gene. They may be tolerated because PCDH15 has several alternate cytoplasmic domain exons and differentially spliced isoforms may function redundantly. Effects of some PCDH15 truncating variants were addressed by deep sequencing of a panethnic population.The Journal of Molecular Diagnostics. 11/2014; 16(6):673–678.
Cadherin 23 and protocadherin 15 interact to form
tip-link filaments in sensory hair cells
Piotr Kazmierczak1,2*, Hirofumi Sakaguchi3*, Joshua Tokita3, Elizabeth M. Wilson-Kubalek1, Ronald A. Milligan1,
Ulrich Mu ¨ller1,2* & Bechara Kachar3*
Hair cells of the inner ear are mechanosensors that transduce
mechanical forces arising from sound waves and head movement
into electrochemical signals to provide our sense of hearing and
balance. Each hair cell contains at the apical surface a bundle of
tips of stereocilia in proximity to extracellular tip-link filaments
that connect the stereocilia and are thought to gate the mechan-
oelectrical transduction channel1–3. Recent reports on the compo-
sition4–8, properties and function9–11of tip links are conflicting29.
Here we demonstrate that two cadherins that are linked to inher-
ited forms of deafness in humans12–15interact to form tip links.
Immunohistochemical studies using rodent hair cells show that
cadherin 23 (CDH23) and protocadherin 15 (PCDH15) localize to
the upper and lower part of tip links, respectively. The amino
termini of the two cadherins co-localize on tip-link filaments.
Biochemical experiments show that CDH23 homodimers interact
in trans with PCDH15 homodimers to form a filament with struc-
tural similarity to tip links. Ions that affect tip-link integrity and a
mutation in PCDH15 that causes a recessive form of deafness16
disrupt interactions between CDH23 and PCDH15. Our studies
define the molecular composition of tip links and provide a con-
ceptual base for exploring the mechanisms of sensory impairment
associated with mutations in CDH23 and PCDH15.
Ultrastructural studies show that tip links are helical filaments
that separate into multiple strands at both ends9,11. Tip links share
properties with cadherins17,18, and CDH23 and PCDH15 have been
localized to tip links4,7. However, several laboratories failed to detect
CDH23 expression attip links5,6, andan antibody to the extracellular
we generated antibodies against CDH23 and PCDH15. Antibodies
EC2 and EC15/EC16 of CDH23, respectively. Antibody PB811 was
raised against a peptide sequence in EC1 of PCDH15. In western
blots, PB264 and PB240 specifically recognized CDH23, whereas
PB811 recognized PCDH15 (Supplementary Fig. 1). In agreement
with earlier findings4–7, immunolocalization experiments revealed
expression of CDH23 and PCDH15 associated with kinociliary links
and transient lateral links in developing hair cells of mice, rats and
stained tip links in mature hair cells (Fig. 1a–e; see also Supplemen-
tary Fig. 2). CDH23 labelling appeared as fluorescent puncta at the
tip-link region (Fig. 1a-e). In splayed stereocilia, the puncta predo-
minantly partitioned with the taller stereocilia (Fig. 1b, c). Removal
stereociliary bundle17. Similarly, when samples were pre-incubated
10min at 25 uC or at 37uC, the CDH23 fluorescent puncta separated
appeared closer to the stereocilia tips, suggesting temperature-
dependent movement and further separation of the upper portion
of the tip link after BAPTA treatment. These findings are consistent
with CDH23 localizing to the upper part of tip links.
immunogold transmission electron microscopy (TEM) to analyse
the localization of PCDH15 and CDH23. Interactions between clas-
sical cadherins such as N- and E-cadherin depend on EC1 (ref. 18).
Each EC domain has a dimension of ,4.5nm19. Assuming a similar
mode of interaction and dimensions for the EC domains in CDH23
and PCDH15, PB811 is expected to bind at ,49.5nm (11 ECs) from
the lower end of tip links, PB264 at ,54nm (12 ECs) and PB240 at
,117nm (26 ECs) (Fig. 1j). Using PB811, PB264 and PB240, we
observed gold labelling at 37617nm (n5113), 52.5619nm
(n5111) and 138634nm (n552) above the lower tip-link inser-
tion site, respectively (Fig. 1f–i). These values agree with the pre-
dicted values and suggest that CDH23 and PCDH15 molecules
interacting at their N termini form tip links (Fig. 1k). However, we
and PCDH15 molecules form the tip link, and some gold particles
associated with CDH23 were further apart from the lower insertion
point of the tip link than predicted, which could be a consequence of
tip-link breakageduringsample preparation. Because itisdifficult to
visualize tip links in thin sections, we confirmed by freeze-etching
immunogold labelling that the CDH23 antibody stained the tip-link
filament (Supplementary Fig. 3).
To determine whether CDH23 and PCDH15 molecules resemble
tip links, we expressed, by transfection in HEK293 cells, the extra-
cellular domains of CDH23 and PCDH15 fused to His affinity tags
(Fig. 2a). The fusion proteins were purified with nickel (Ni-NTA)
beads (Fig. 2b) and analysed by negative staining TEM. In the pres-
ence of 1mM Ca21, CDH23–His molecules formed dimeric fila-
ments with helical appearance (Fig. 2c and Supplementary Fig. 4).
Frequently the two strands splayed at one filament end and formed a
branched or looped structure. The extracellular domain of CDH23
consists of 27 EC domains with an estimated size of ,122nm, which
is in good agreement with the observed length of 12964.5 nm. In
the absence of Ca21, the CDH23 strands lost their filamentous
shape, suggesting Ca21-dependent folding of CDH23 (Supplemen-
tary Fig. 4). The extracellular domain of PCDH15 also formed inter-
twined dimers (Fig. 2e and Supplementary Fig. 5) with a length of
*These authors contributed equally to this work.
1The Scripps Research, Institute Department of Cell Biology,2Institute for Childhood and Neglected Disease, La Jolla, California 92037, USA.3Laboratory of Cellular Biology, National
Institute on Deafness and other Communication Disorders, National Institutes of Health, Bethesda, Maryland 20892, USA.
Vol 449|6 September 2007|doi:10.1038/nature06091
51.962.7nm, which is in agreement with the ,49.5nm predicted
To define the orientation of extracellular domains of CDH23 and
PCDH15 in homodimers, we used Ni-NTA beads coupled to nano-
gold particles (Fig. 2d, f and Supplementary Fig. 5). Several Ni-NTA
beads can bind to each His tag20. For CDH23, we analysed 195 fila-
gold particles at one end, whereas only 3 showed nanogold particles
at both ends (Fig. 2d, f and Supplementary Fig. 5). For PCDH15, we
analysed 58 filaments. A total of 39 were labelled at one end and the
remaining filaments were not labelled. In CDH23, the gold beads
attached to the splayed end of the filaments (Fig. 2d), suggesting that
the branch point is membrane proximal. These findings suggest that
the extracellular domains of CDH23 and PCDH15 form preferen-
tially parallel homodimers.
Thelocalization of CDH23 andPCDH15 at opposite tip-link ends
suggests that they interact to form tip links. To analyse interactions,
we generated constructs containing the extracellular domain of
heavy chain (Fig. 3a). We independently expressed CDH23–His,
CDH23–Fc, PCDH15–His and PCDH15–Fc, mixed the proteins in
the presence of 1mM Ca21, and isolated protein complexes using
Ni-NTA beads, which bind to the His tag but not the Fc tag. Western
blot analysis revealed that PCDH15–Fc, but not CDH23–Fc, inter-
acted with CDH23–His (Fig. 3b). PCDH15–His interacted with
CDH23–Fc but not with PCDH15–Fc (Fig. 3c). We conclude that
CDH23 and PCDH15 engage in heterophilic but not homophilic
interactions. We were surprised that CDH23–His and CDH23–Fc
CDH23 (and PCDH15) molecules appeared in electron micrograph
images as dimers (Fig. 2 and Supplementary Figs 4 and 5), suggesting
that the proteins are purified as stable dimers. Furthermore, whereas
CDH23 can mediate homophilic binding7, the washing conditions
used here were of greater stringency than in previous studies and
suggest that homophilic binding is of low affinity. Homophilic adhe-
sion could be important in cells expressing high CDH23 levels.
Adhesion between classical cadherins is optimal at$1mM Ca21
(ref. 18), whereas tip links are bathed in the sub-millimolar Ca21
PCDH15 were disrupted in the absence of Ca21, but were maximal
at$0.1mM Ca21, which is within the range of the Ca21concentra-
tion of the endolymph21. Tip links are disrupted by treatment with
LaCl3, but not by treatment with BaCl2(ref. 9). Similarly, complex
~52 nm ~130 nm
Figure 1 | Tip links are formed by CDH23 and PCDH15. a–e, CDH23
antibodies PB240 (a, b, d, e) and PB264 (c) labelled (green) the tip-link
was visualized with rhodamine–phalloidin (red). In splayed stereocilia
(b, c) CDH23 fluorescent puncta frequently localized at the lateral wall of
stereocilia (arrowheads). After treatment of samples with BAPTA and
At 37uC, fluorescent puncta moved towards the stereociliary tip.
f–h, Immunogold electron microscopic images of the tip-link region of
guinea-pig outer hair cells. i, Distances of gold particles from the lower
insertion site of tip links, demonstrating overlap of labelling for PB811 and
PB264. j, Diagram of CDH23 and PCDH15 showing the localization of the
tip links. Scale bars: a–e, 2.5mm; f–h, 50nm.
NATURE|Vol 449|6 September 2007
formation between CDH23 and PCDH15 in the presence of 1mM
Ca21was inhibited by LaCl3but not by BaCl2(Fig. 3d).
To determine which EC domains promote interactions between
CDH23 and PCDH15, we generated fusion proteins containing the
did not generate constructs with one or two EC domains because
low binding affinity22. The 11 or 3 EC domains of CDH23 interacted
with PCDH15 (Fig. 3e). To verify interactions independently, we
expressed the transmembrane version of PCDH15 in NIH3T3 cells
and incubated non-permeabilized cells with CDH23–Fc (Fig. 4a).
The cells were then permeabilized; PCDH15 was visualized with an
antibody to the cytoplasmic domain, and CDH23–Fc with an Fc-
specific antibody. CDH23–Fc bound cells expressing PCDH15 but
notcontrol cells.Thepattern ofCDH23–Fcbinding overlapped with
PCDH15 staining at the cell membrane. Intracellular vesicles con-
taining PCDH15 did not bind CDH23–Fc because CDH23–Fc was
added before permeabilization. Binding was also observed with the
omitted (Fig. 4a).
To determine which PCDH15 domains mediate interactions with
CDH23, we generated constructs consisting of the N-terminal 3 EC
domains of PCDH15, but the proteins were unstable. Single amino
acid changes in PCDH15 are linked to non-syndromic recessive
deafness DFNB23 (ref. 16). We engineered two DFNB23-associated
mutations into PCDH15–Fc (Fig. 4b). A mutation in EC1, but not
EC2, abolished interactions between PCDH15–Fc and CDH23–His
(Fig. 4c). We conclude that EC1 of PCDH15 is critical for adhesion,
and that DFNB23 is caused in part by adhesion defects. Mutations
such as those in EC2 may cause deafness by affecting the mechanical
properties of tip links.
We next analysed a mix of CDH23–His and PCDH15–His by
electron microscopy. Most molecules appeared as cis-homodimers,
but we consistently observed complexes with the proper dimensions
for CDH23 homodimers interacting in trans with PCDH15 homo-
dimers (Fig. 4d; see also Supplementary Fig. 6) to form a complex
,180nm in length, which is in good agreement with the reported
length of tip links (150–200nm)9,11. One end of the complex fre-
quently branched, which we infer to be the end where CDH23 is
located. These findings provide further evidence that CDH23 and
PCDH15 interact at their N termini.
Collectively, our findings provide evidence that tip links areasym-
metric adhesion complexes consisting of CDH23 and PCDH15
interacting at their N termini and forming the upper and lower
part of tip links, respectively. CDH23 and PCDH15 interact at
0.1mM Ca21, which is below the concentration necessary for
in the EC1 domain of classical cadherins19,24are not conserved in
CDH23, and only one Trp residue is present in EC1 of PCDH15,
suggesting that the binding surface between CDH23 and PCDH15
is optimized for endolymph conditions. Our findings also suggest
that the two ends of a tip link are functionally distinct. Myo1c, the
adaptation motor for the mechanoelectrical transduction channel,
co-immunoprecipitated with CDH23 (ref. 7) and is absent from
stereocilia of CDH23-deficient waltzer mice25, suggesting that
Myo1c interacts with the upper tip-link end. The membrane at the
lower tip-link end appears under tension and is pulled upwards9,
raising thepossibility that part of thegating spring may localize close
to PCDH15. The localization of CDH23 and PCDH15 at kinociliary
and transient lateral links in hair cells (Supplementary Fig. 2h, i)
actions betweenthetwocadherins. Thismolecular asymmetrymight
be important for translating extracellular signals into cytoskeletal
changes that determine the polarity of the stereociliary bundles.
Consistent with this model, bundle polarity is affected in mice with
a mutation in Pcdh15 (ref. 26).
No goldNTA gold
Figure 2 | The extracellular CDH23 and PCDH15 domains form parallel
homodimers. a, Diagram of CDH23 and PCDH15 constructs. The
HEK293 cells. Full-length proteins were detected by western blot analysis
and silver staining. No protein was visible in the supernatant of
c–f, Analysis of CDH23–His and PCDH15–His by negative staining
transmission electron microscopy. Strands are traced in the right panels for
at one end. d, Ni-NTA nanogold beads (arrow) attached to the end of
CDH23 homodimers that formed the branch/loop. Inset: sizes of a single
and multiple nanogold beads (fivefold enlarged). In the right panels, gold
beads are shown as black dots. e, PCDH15–His molecules formed dimers.
f, Ni-NTA nanogold beads (arrow) attached to one end of the PCDH15–His
homodimers. Scale bars: c, d, 100nm; e, f, 50nm.
NATURE|Vol 449|6 September 2007
Immunolocalization. Antibodies to CDH23 and PCDH15 were developed
in rabbits and affinity purified. Antibodies PB264 and PB240 were raised
against aminoacids123–135(GDVNDNAPTFHNQ) and
(ATRPAPPDRERQ) of murine CDH23 (NM_023370), respectively. PB811
was raised against amino acids 80–96 (LSLKDNVDYWVLLDPVK) of
PCDH15 (NP_075604). Immunofluorescence microscopy was performed with
labelling, tissues from guinea-pigs (P90–P120) were fixed and labelled with
goat anti-rabbit IgG (BB International). Freeze-etching following immunogold
labelling was carried out as described9.
Biochemistry. Expression constructs were generated by polymerase chain
reaction (PCR) using described constructs7,26and methods28and introduced
in pCDNA3 (Invitrogen). The Fc domain was derived from hIgG1DNApRK
(Genentech). HEK293 cells were transfected using FuGENE6 reagent
(Roche), medium was collected and His-tagged proteins purified by affinity
chromatography using Ni-NTA beads (Qiagen). Fc fusions were produced in
a similar way. For structural studies, CDH23–His and PCDH15–His were
applied to carbon grids, stained with 1% uranyl acetate and imaged using a
with Fc-tagged PCDH15, protein complexes were purified with Ni-NTA beads
and analysed by western blotting. CDH23 was detected with PB240, Fc-tagged
proteins with horseradish peroxidase (HRP)-conjugated anti-human IgG
(Amersham Biosciences), and PCDH15–His with HRP-conjugated anti-Myc
antibody (Upstate Biotechnology). For immunocytochemistry, NIH3T3 cells
were transfected with full-length PCDH15 (ref. 28), incubated with medium
containing Fc-tagged CDH23, washed, fixed and stained with Alexa-488-
conjugated anti-IgG antibody (Molecular Probes). Cells were permeabilized
and re-stained with anti-PCDH15 cytodomain antibody28, and with DAPI
CDH23–His pull down
133 VIYHEVRIVV 142
259 DVLDGDDLGP 268
Figure 4 | Analysis of interactions between CDH23 and PCDH15. a, Full-
length PCDH15 was expressed by transfection in NIH3T3 cells. Cells were
incubated with recombinant CDH23–Fc and CDH23F11–Fc (indicated
on the top of the panels). PCDH15 (red) and CDH23 (green) were
shows DAPI staining (blue). Only cells expressing PCDH15 bound
CDH23–Fc and CDH23F11–Fc. No staining wasobserved when CDH23–Fc
was omitted. b, PCDH15–Fc constructs with point mutations in EC1 and
EC2 found in DFNB23 patients16. c, CDH23–His was incubated with the
indicated proteins, complexes were purified and analysed by western blot
analysis. Top panel: CDH23–His binds to PCDH15–Fc and
d, CDH23–His and PCDH15–His were mixed and analysed by negative
staining and electron microscopy. CDH23–His dimers, black asterisk;
PCDH15–His dimers, white asterisk; CDH23 and PCDH15 contact zone,
white arrows; branched N-terminus of CDH23, black arrows. Scale bars:
a, 20mm; d, 50nm.
CDH23–His pull down
PCDH15–His pull down
PCDH15–His pull down
1 mM CaCl2
0 0.1 1 5 10 (mM)
0 0.1 1 5 10 (mM)
CDH23–Fc ControlCDH23–Fc Control
Figure 3 | CDH23 binds to PCDH15. a, Diagram of CDH23 and PCDH15
constructs. The proteins were purified from transfected HEK293 cells.
b, CDH23–His was incubated with CDH23–Fc or PCDH15–Fc, protein
complexes were purified and analysed by western blot analysis. In control
lanes, CDH23–Hiswas omitted. Toppanel: CDH23–His bindsPCDH15–Fc
loading controls for the Fc fusion proteins and CDH23 (detected with an
antibody to Fc and CDH23, respectively). c, PCDH15–His was incubated
with CDH23–Fc or PCDH15–Fc. In control lanes, PCDH15–His was
omitted. The analysis of the protein complexes was carried out as described
above. PCDH15–His bound to CDH23–Fc but not to PCDH15–Fc. d, Top
panel: CDH23–His was incubated with PCDH15–Fc. Samples were washed
in buffer containing the indicated amounts of CaCl2. Complex formation
was observed in the presence of$0.1mM CaCl2. Bottom panels:
and the indicated amounts of LaCl3or BaCl2. Complex formation was
disrupted only by LaCl3. e, PCDH15–His was incubated with the indicated
proteins; complexes were purified and analysed by western blotting.
PCDH15–His bound to CDH23F11–Fc (N-terminal 11 EC domains) and
CDH23F3–Fc (3 EC domains).
NATURE|Vol 449|6 September 2007
Full Methods and any associated references are available in the online version of
the paper at www.nature.com/nature.
Received 17 April; accepted 10 July 2007.
Corey, D. Sensory transduction in the ear. J. Cell Sci. 116, 1–3 (2003).
Gillespie, P. G. & Walker, R. G. Molecular basis of mechanosensory transduction.
Nature 413, 194–202 (2001).
Hudspeth, A. J. How hearing happens. Neuron 19, 947–950 (1997).
complex of sensory hair cells, is protocadherin-15. J. Neurosci. 26, 7022–7034
Lagziel, A. et al. Spatiotemporal pattern and isoforms of cadherin 23 in wild type
and waltzer mice during inner ear hair cell development. Dev. Biol. 280, 295–306
Michel, V. et al. Cadherin 23 is a component of the transient lateral links in
the developing hair bundles of cochlear sensory cells. Dev. Biol. 280, 281–294
Siemens, J. et al. Cadherin 23 is a component of the tip link in hair-cell stereocilia.
Nature 428, 950–955 (2004).
Sollner, C. et al. Mutations in cadherin 23 affect tip links in zebrafish sensory hair
cells. Nature 428, 955–959 (2004).
Kachar, B., Parakkal, M., Kurc, M., Zhao, Y. & Gillespie, P. G.
High-resolution structure of hair-cell tip links. Proc. Natl Acad. Sci. USA 97,
10. Sotomayor, M., Corey, D. P. & Schulten, K. In search of the hair-cell gating
spring elastic properties of ankyrin and cadherin repeats. Structure 13, 669–682
11.Tsuprun, V., Goodyear, R. J. & Richardson, G. P. The structure of tip links
and kinocilial links in avian sensory hair bundles. Biophys. J. 87, 4106–4112
12. Ahmed, Z. M. et al. Mutations of the protocadherin gene PCDH15 cause Usher
syndrome type 1F. Am. J. Hum. Genet. 69, 25–34 (2001).
13. Bolz, H. et al. Mutation of CDH23, encoding a new member of the cadherin gene
family, causes Usher syndrome type 1D. Nature Genet. 27, 108–112 (2001).
14. Bork, J. M. et al. Usher syndrome 1D and nonsyndromic autosomal recessive
deafness DFNB12 are caused by allelic mutations of the novel cadherin-like gene
CDH23. Am. J. Hum. Genet. 68, 26–37 (2001).
15. Di Palma, F. et al. Mutations in Cdh23, encoding a new type of cadherin, cause
stereocilia disorganization in waltzer, the mouse model for Usher syndrome type
1D. Nature Genet. 27, 103–107 (2001).
Genet. 12, 3215–3223 (2003).
17. Assad, J. A., Shepherd, G. M. & Corey, D. P. Tip-link integrity and mechanical
transduction in vertebrate hair cells. Neuron 7, 985–994 (1991).
18. Halbleib, J. M. & Nelson, W. J. Cadherins in development: cell adhesion, sorting,
and tissue morphogenesis. Genes Dev. 20, 3199–3214 (2006).
19. Boggon, T. J. et al. C-cadherin ectodomain structure and implications for cell
adhesion mechanisms. Science 296, 1308–1313 (2002).
20. Buchel, C., Morris, E., Orlova, E. & Barber, J. Localisation of the PsbH subunit in
photosystem II: a new approach using labelling of His-tags with a Ni21-NTA gold
cluster and single particle analysis. J. Mol. Biol. 312, 371–379 (2001).
21. Anniko, M. & Wroblewski, R. Ionic environment of cochlear hair cells. Hear. Res.
22, 279–293 (1986).
22. Brieher, W. M., Yap, A. S. & Gumbiner, B. M. Lateral dimerization is required for
the homophilic binding activity of C-cadherin. J. Cell Biol. 135, 487–496 (1996).
recombinant extracellular domain of E-cadherin upon calcium binding. Eur. J.
Biochem. 223, 1019–1026 (1994).
24. Patel,S.D.etal.TypeIIcadherinectodomain structures:implicationsforclassical
cadherin specificity. Cell 124, 1255–1268 (2006).
25. Phillips, K. R., Tong, S., Goodyear, R., Richardson, G. P. & Cyr, J. L. Stereociliary
myosin-1c receptors are sensitive to calcium chelation and absent from cadherin
23 mutant mice. J. Neurosci. 26, 10777–10788 (2006).
26. Senften, M. et al. Physical and functional interaction between protocadherin 15
and myosin VIIa in mechanosensory hair cells. J. Neurosci. 26, 2060–2071
27. Schneider, M. E. et al. A new compartment at stereocilia tips defined by spatial
and temporal patterns of myosin IIIa expression. J. Neurosci. 26, 10243–10252
28. Siemens, J. et al. The Usher syndrome proteins cadherin 23 and harmonin form a
complex by means of PDZ-domain interactions. Proc. Natl Acad. Sci. USA 99,
channel, and gating spring of the hair cell? Curr. Opin. Neurobiol. 15, 389–396
Supplementary Information is linked to the online version of the paper at
Acknowledgements We thank M. Sotomayor for discussions regarding cadherin
ectodomain structure. We thank members of the Mu ¨ller and Kachar laboratories
for comments. This work was supported by the NIH (U.M., R.A.M., E.M.W.-K., and
H.S., J.T., B.K. (intramural funding)). Negative staining TEM analysis was
conducted at the National Resource of Automated Molecular Microscopy.
as co-senior authors. P.K and H.S. are co-first authors. H.S. and J.T. characterized
and 3. P.K. carried out the experiments in Figs 2–4 and Supplementary Figs 1, 4, 5
and 6. P.K. provided the proteins for negative staining TEM, which was carried out
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial interests.
Correspondence and requests for materials should be addressed to U.M.
(firstname.lastname@example.org) or B.K. (email@example.com).
NATURE|Vol 449|6 September 2007
Antibodies and immunofluorescence. Anti-peptide antibodies specific to the
longest known isoforms of CDH23 (PB240 and PB264) and PCDH15 (PB811)
were developed in rabbits and affinity-purified. The polypeptides matched
regions in the extracellular domain of the longest isoforms of the mouse
CDH23 and PCDH15 proteins. Antibodies PB264 and PB240 were raised
against aminoacids 123–135(GDVNDNAPTFHNQ) and
(ATRPAPPDRERQ) of CDH23 (NM_023370), respectively. Both sequences
are part of the Ca21-binding motifs of EC domains: DVND at the interface
between EC1 and EC2; DRERQ at the interface between EC15 and EC16.
PB811 was raised against LSLKDNVDYWVLLDPVK corresponding to amino
acids 80–96 in the EC1 of PCDH15 (NP_075604). Inner ear tissues were
obtained from rats, mice and guinea-pigs killed in accordance to National
Institute of Health guidelines under the NIDCD animal protocol 1215-05.
Immunofluorescence microscopy was performed in whole-mount prepa-
rations as reported27. Antibodies were used at ,2mgml21. PB240 and PB264
immunolabelling depended on removal of Ca21from the buffer after fixation,
probably owing to epitope unmasking. Ca21removal was attained by per-
forming the entire labelling procedure using a 5mM BAPTA/HEPES-buffered
Hank’s balanced salt solution (HBHBSS) to dilute the primary and secondary
Immunogold labelling. Tissues were dissected from adult guinea-pigs (P90–
glutaraldehydewith2% tannicacid followedbystainingwith1% uranylacetate.
(Electron Microscopy Science). Semi-thin sections (250– 500-nm thick) were
cut and viewed using a Zeiss 922 transmission electron microscope equipped
tissues processed as described above were further processed as described9. The
images were analysed using NIH Image software.
allowthe solutionto sufficiently substituteendolymph. Sampleswere incubated
in 5mM BAPTA/HBHBSS for 1min at room temperature, and recovered in
HBHBSS with 3mM Ca21for 10min at 25uC or 37uC. Samples were fixed in
4% paraformaldehyde and labelled with PB240 as described above. The images
were analysed using NIH Image software.
Expression and purification of CDH23 and PCDH15. The extracellular
domains of CDH23 and PCDH15 were amplified by PCR using described com-
plementary DNA constructs7,26and methods28. Cloning procedures were as fol-
lows (clones verified by sequencing): to clone murine Cdh23 cDNA
(NM_023370) into pCDNA3, the BglII site in pCDNA3 was removed by cutting
with BglII, filling recessed termini with Klenow polymerase and re-ligation.
The 59 end of the CDH23 cDNA was amplified by PCR (GGGGTACCATGG-
GTACTCCCTGGTCAC; CCGAATTCGGGGTCATTGTCATTGAGATC) and
cloned into KpnI/EcoRI sites of pCDNA3DBglII. The 39 end of the CDH23
pCDNA3DBglII containing the 59 end of the CDH23 cDNA. A 59 BglII/FseI
CDH23 fragment was inserted to yield the final vector. To generate His-tagged
CDH23, double-stranded oligonucleotides encoding the tag were ligated into
EcoRI/XhoI sites of pCDNA3DBglII (AATTCGGGCATCATCATCATCATC-
ATGCCG) to generate pCDNA3– His. To generate Fc-tagged versions, the
human IgG1 Fc fragment was amplified by PCR (GGCGAATTCGGGCC-
from plasmid hIGG1DNApRK (Genentech) and cloned into EcoR1/Xho1 of
pCDNA3 to generate pCDNA3Fc. CDH23 cDNA amplified by PCR (primers:
AGACATGTCAT) was cloned into KpnI/EcoRI of pCDNA3His and pCDNAFc.
a double-stranded oligonucleotide encoding an Myc tag (AGCTTGGGGTA-
TGCCATGAATTCCCCTTCTGTGTACCCCA) was inserted 59 of the His tag
in HindIII/EcoRI sites of pCDNA3–His. The human IgG1 Fc fragment was
amplified by PCR (GGAGAAGCTTGGGGTACACAGAAGGGCCACGAGG-
AGACAAAACTCAC; CGAGCTCGAGTCATTTACCCGGAGACAGGGA) and
cloned into HindIII/XhoI sites of pCDNA3. A HindIII fragment containing
the full PCDH15 extracellular domain was inserted. To generate truncation
mutations and point mutations, the following primers were used (only
primers at the truncation/mutation site are shown; other primers are described
above): CDH23F11–Fc, CGCGAATTCCTGCTGTGTGAATACAGGGGCCT-
TGT; PCDH15(DFNB23-EC1), ATCTATCATGAGTAGCCATCGTGGTGC-
GAGAT; PCDH15(DFNB23-EC2), TGGAGATGACCTGGACCCTATGTTT-
HEK293 cells were grown in DMEM (Invitrogen) containing 10% fetal
bovine serum (Gemini Products), 13 penicillin/streptomycin/glutamine
(Invitrogen). Cells were transfected using FuGENE6 reagent (Roche). Twelve
hours after transfection, medium was replaced with serum-free DMEM/F12
supplemented with antibiotics and glutamine. Forty-eight hours later medium
was collected and subjected to chromatography over Ni-NTA agarose (Qiagen).
His-tagged proteins were eluted with 100mM imidazole in 50mM Tris-HCl,
150mM NaCl, pH7.5. For interaction studies, CDH23 and PCDH15 tagged
with human IgG1 Fc were produced in a similar way. Purified proteins were
separated on 7.5% SDS–PAGE gels and analysed by silver staining using a kit
and PCDH15–His were applied to glow-discharged carbon grids and stained
with 1% uranyl acetate. Images were acquired on a Philips CM200 FEG trans-
F20 with a 4K by 4K Gatan CCD camera.
and 25ml Ni-NTA agarose beads (Qiagen) for 2h at 25uC on a rotating wheel.
150mM NaCl, 1mM CaCl2, pH7.5. Beads were re-suspended in sample buffer
and analysed by SDS–PAGE (7.5% gels) and western blot analysis. Anti-CDH23
antibody PB240 was used to detect His-tagged CDH23, HRP-conjugated anti-
human IgG (Amersham Biosciences) to detect Fc-tagged proteins, and HRP-
conjugated anti-Myc antibody (Upstate Biotechnology) to reveal PCDH15
Immunocytochemistry. NIH3T3 cells were cultured on glass coverslips and
4uC, washed, fixed for 10min in 2% paraformaldehyde and stained with Alexa-
488-conjugated anti-human IgG antibody (Molecular Probes). Cells were per-
meabilized with 0.1% Triton and re-stained with anti-PCDH15 cytodomain
antibody28and with DAPI (Molecular Probes) to reveal nuclei.