Article
TMC1 and TMC2 Localize at the Site of
Mechanotransduction in Mammalian Inner Ear Hair
Cell Stereocilia
Graphical Abstract
Highlights
d TMC1 and TMC2 localize at the site of hair cell
mechanotransduction
d Expression of TMC2 in cochlea is limited to developing
stereocilia bundles
d TMC1 is an essential component of the mechanotransduction
channel complex
Authors
Kiyoto Kurima, Seham Ebrahim,
Bifeng Pan, ..., Jeffrey R. Holt,
Andrew J. Grifﬁth, Bechara Kachar
Correspondence
grifﬁta@nidcd.nih.gov (A.J.G.),
kacharb@nidcd.nih.gov (B.K.)
In Brief
The molecular makeup of hair cell
mechanoelectrical transduction (MET)
complex remains elusive. Kurima et al.
show that ﬂuorophore-tagged TMC1/2
localize to MET sites at stereocilia tips
and that their expression in TMC1/2-null
mice rescues MET. These ﬁndings
support the model that TMC1/2 are part of
the MET complex.
Kurima et al., 2015, Cell Reports 12, 1–12
September 8, 2015 ª2015 The Authors
http://dx.doi.org/10.1016/j.celrep.2015.07.058
Cell Reports
Article
TMC1 and TMC2 Localize at the Site
of Mechanotransduction in Mammalian
Inner Ear Hair Cell Stereocilia
Kiyoto Kurima,1,5 Seham Ebrahim,2,5 Bifeng Pan,3 Miloslav Sedlacek,2 Prabuddha Sengupta,4 Bryan A. Millis,2
Runjia Cui,2 Hiroshi Nakanishi,1 Taro Fujikawa,2 Yoshiyuki Kawashima,1 Byung Yoon Choi,1 Kelly Monahan,1
Jeffrey R. Holt,3 Andrew J. Grifﬁth,1,* and Bechara Kachar2,*
1Molecular Biology and Genetics Section
2Laboratory of Cell Structure and Dynamics
National Institute on Deafness and Other Communication Disorders, NIH, Bethesda, MD 20892, USA,
3Department of Otolaryngology, F.M. Kirby Neurobiology Center, Boston Children’s Hospital, Harvard Medical School, Boston,
MA 02115, USA
4Cell Biology and Metabolism Program, Eunice Kennedy Shriver National Institute of Child Health and Human Development, Bethesda,
MD 20892, USA
5Co-ﬁrst author
*Correspondence: grifﬁta@nidcd.nih.gov (A.J.G.), kacharb@nidcd.nih.gov (B.K.)
http://dx.doi.org/10.1016/j.celrep.2015.07.058
This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
SUMMARY
Mechanosensitive ion channels at stereocilia tips
mediate mechanoelectrical transduction (MET) in inner
ear sensory hair cells. Transmembrane channellike
1 and 2 (TMC1 and TMC2) are essential for MET
and are hypothesized to be components of the
MET complex, but evidence for their predicted
spatiotemporal localization in stereocilia is lacking.
Here, we determine the stereocilia localization of
the TMC proteins in mice expressing TMC1-mCherry
and TMC2-AcGFP. Functionality of the tagged proteins
was veriﬁed by transgenic rescue of MET
currents and hearing in Tmc1D/D
;Tmc2D/D
mice.
TMC1-mCherry and TMC2-AcGFP localize along
the length of immature stereocilia. However, as hair
cells develop, the two proteins localize predominantly
to stereocilia tips. Both TMCs are absent
from the tips of the tallest stereocilia, where MET
activity is not detectable. This distribution was
conﬁrmed for the endogenous proteins by immunoﬂuorescence.
These data are consistent with TMC1
and TMC2 being components of the stereocilia
MET channel complex.
INTRODUCTION
Mechanoelectrical transduction (MET), whereby mechanical
stimuli are converted to electrical signals, is an integral property
of inner ear hair cells, accomplished by their mechanosensory
organelle, the stereocilia hair bundle. Each bundle comprises
dozens of actin-based protrusions with graded lengths, organized
in a staircase array. Tip links, extracellular protein ﬁlaments
composed of cadherin-23 (CDH23) and protocadherin-15
(PCDH15), connect pairs of adjacent stereocilia near their tips
in the direction of optimal mechanosensitivity of the bundle (Kazmierczak
et al., 2007; Sakaguchi et al., 2009). At each end of the
tip link are densely packed macromolecular complexes underlying
the stereocilia membrane, known as tip-link insertion plaques.
The upper insertion plaque is presumed to contain a
cluster of motor proteins (Grati and Kachar, 2011) that maintains
a resting tension on the tip link (Schwander et al., 2010). The
lower tip-link insertion site is thought to be the site for the MET
channel complex (Beurg et al., 2009). Stereocilia-mediated
MET occurs as a consequence of stereocilia bundle deﬂection
toward the tallest row of stereocilia, which exerts tension on tip
links, opening MET channels and causing depolarization of the
hair cell. The developmental acquisition of MET, which has
been spatiotemporally characterized in rat (Waguespack et al.,
2007) and mouse (Lelli et al., 2009) is tonotopic, with onset of
MET in hair cells in the organ of Corti between postnatal day
(P) 1 and P2 and fully mature MET being reached by P8.
Aside from the tip-link proteins, the molecular composition of
the MET apparatus remains elusive (Fuchs, 2015; Gillespie et al.,
2005). The tetraspanin TMHS/LHFPL5 (Beurg et al., 2015; Xiong
et al., 2012) and a protein with two transmembrane domains,
TMIE (Zhao et al., 2014), have recently been identiﬁed as MET
channel accessory components. However, the precise spatiotemporal
localization of these proteins and the identities of the
pore-forming units and other essential auxiliary elements are
yet to be determined. Two members of the transmembranechannel-like
(TMC) family, TMC1 and TMC2, are candidates to
be part of the MET complex based on several lines of evidence:
(1) Tmc1 and Tmc2 mRNA are selectively expressed in developing
hair cells at the onset of acquisition of mechanosensitivity
(Kawashima et al., 2011); (2) mutations of TMC1 cause deafness
in humans and mice (Kurima et al., 2002; Vreugde et al., 2002);
(3) in the absence of both functional TMC1 and TMC2, hair cells
of the mouse auditory and vestibular system lack mechanosensory
responses to forward deﬂection of the hair bundle (Beurg
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 1
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
et al., 2014; Kawashima et al., 2011); (4) TMC1- and TMC2deﬁcient
hair cells fail to take up FM1-43 or gentamicin (Kawashima
et al., 2011), which enter wild-type hair cells via the MET
channel (Gale et al., 2001; Marcotti et al., 2005); (5) transient
exogenous expression of either TMC1 or TMC2 restored mechanosensitivity
in hair cells from double homozygous null mice
(Tmc1D/D
;Tmc2D/D
), demonstrating that these proteins are
necessary for generating MET currents (Kawashima et al.,
2011); (6) ﬁsh and mouse orthologs of the cytoplasmic domain
of PCDH15 were recently shown to interact with both TMC1
and TMC2 (Maeda et al., 2014); and (7) a more-recent study
shows interactions between PCDH15 as well as the MET protein
LHFPL5 with TMC1 and TMC2 (Beurg et al., 2015).
Whereas these data collectively suggest that TMC1 and TMC2
are crucial for the development and gating of MET currents (Kawashima
et al., 2015), to ascertain whether they are indeed
components of the MET complex, it is necessary to verify their
localization at the site of the MET channel complex, at the lower
tip-link insertion sites on shorter stereocilia, from the early postnatal
period onward (Beurg et al., 2009). Previous studies reported
targeting of ﬂuorophore-tagged TMC1 and TMC2 to the
apex of all stereocilia when exogenously expressed using biolistic
gene-gun mediated transfection of hair cells cultured
in vitro (Kawashima et al., 2011), and Beurg et al. (2015) localized
TMC1 to stereocilia and kinocilia using antibody labeling.
However, a deﬁnitive conclusion on the precise spatiotemporal
localization of the proteins during postnatal development was
precluded by the transient nature of the protein expression, the
likelihood of overexpression due to the non-native promoter
used, and potential for cell damage by the gene-gun method.
We addressed this in the current study by generating and characterizing
transgenic mice that express TMC1-mCherry and
TMC2-AcGFP under their native promoters. We ﬁrst veriﬁed
that the ﬂuorophore-tagged TMC proteins mimic the function
of the native TMC proteins by conﬁrming rescue of hearing and
vestibular deﬁcits, as well as hair cell MET, in Tmc1D/D
;Tmc2D/D
mice expressing TMC1-mCherry and TMC2-AcGFP. Using highperformance
confocal microscopy, we show that both TMC1mCherry
and TMC2-AcGFP localize to the presumed site of
MET at stereocilia tips. The pattern of localization of the ﬂuorophore-tagged
TMC proteins in the transgenic mice was further
validated by immunoﬂuorescence localization of the native proteins
using speciﬁc anti-TMC1 and -TMC2 antibodies. Finally,
we also examined the relationship of TMC1 and TMC2 to the
recently described non-canonical, reverse-polarity mechanosensitive
currents elicited after disruption of conventional MET
(Barr-Gillespie and Nicolson, 2013; Kim et al., 2013; Marcotti
et al., 2014).
RESULTS
Generation of TMC1 and TMC2 BAC Transgenic Mice
To visualize TMC proteins in hair cells, we generated transgenic
mice expressing TMC1 or TMC2 fused to ﬂuorophore tags at
their C termini. The transgenes were derived from bacterial artiﬁcial
chromosomes (BACs) encoding mouse genomic Tmc1 or
Tmc2 to mimic endogenous expression. The native stop codons
were replaced with cDNA encoding mCherry or AcGFP, respectively
(Figure 1A). We obtained four founder lines segregating
Tmc1-mCherry and four founder lines segregating Tmc2AcGFP.
Using qPCR, we estimated that the transgene copy
numbers for Tmc1-mCherry lines 1, 2, 3, and 4 were 3, 7, 7,
and 18, respectively; transgene copy numbers for Tmc2-AcGFP
lines 1, 2, 3, and 4 were 2, 3, 3, and 5, respectively. To determine
whether the TMC1-mCherry and TMC2-AcGFP were functional
and could substitute for the function of native TMC1 and TMC2
in vivo, each transgenic founder line was crossed with the previously
described null Tmc1D/D
, Tmc2 D/D
, and Tmc1D/D
;Tmc2D/D
lines (Kawashima et al., 2011). All transgenic mice were fertile
on the Tmc1D/D
;Tmc2D/D
background and transmitted the transgenes
to offspring. In Tmc1-mCherry line 3, however, transmission
of the transgene through males was restricted to female
offspring, indicating X chromosome integration of the transgene
in this line.
Functional Evaluation of TMC1-mCherry and
TMC2-AcGFP
Fluorophore tags may alter the conﬁguration of proteins to which
they are appended. To evaluate the functionality of TMC1mCherry,
we recorded auditory brainstem response (ABR)
threshold measurements from transgenic mice from founder
lines 1, 2, or 4 segregating Tmc1-mCherry on a Tmc1D/D
background.
Tmc1D/D
mice are profoundly deaf, but we determined
that hearing in these mice was fully rescued when Tmc1mCherry
was present (Figure 1B), indicating that the mCherrytagged
TMC1 can replace the native protein without detectable
changes in auditory function. No difference was observed in the
ABR thresholds among the different Tmc1-mCherry founder
lines despite differences in transgene copy number, suggesting
that the targeting and function of TMC1 are not signiﬁcantly
affected by gene dosage. Only partial rescue of auditory function
was observed in females from line 3 (Figure 1B), where the Tmc1mCherry
transgene segregated in a pattern consistent with
X-linked inheritance (Lyon, 1961). The presence of Tmc2-AcGFP
did not rescue hearing in Tmc1D/D
mice (Figure 1B), consistent
with previous reports on the inability of TMC2 to compensate
for lack of TMC1 in the mature auditory system (Kawashima
et al., 2011). Finally, the abnormal motor vestibular phenotypes,
including head bobbing and arching, present in Tmc1D/D
;
Tmc2D/D
mice (Kawashima et al., 2011), were not detected
when either Tmc1-mCherry, Tmc2-AcGFP, or both transgenes
were present (data not shown). This rescue of function indicates
that the ﬂuorophore tags did not signiﬁcantly affect the function
of TMC1 or TMC2 in the vestibular system and that either TMC1
or TMC2 are sufﬁcient to restore function in the vestibular
system.
We next investigated the effects on whole-cell MET currents
when native TMC1 or TMC2 were replaced with ﬂuorophoretagged
TMC1 or TMC2. As previously reported, inner hair cells
(IHCs) and outer hair cells (OHCs) from Tmc1D/D
;Tmc2D/D
mice
have no detectable MET currents (Kawashima et al., 2011). We
recorded whole-cell MET currents from IHCs and OHCs
of Tmc1D/D
;Tmc2D/D
mice expressing either Tmc1-mCherry
(line 2), Tmc2-AcGFP (line 3), or both transgenes, in developing
(P3) and mature (P7) stereocilia (Figure 1C). At P3, we observed
partial rescue of MET currents in both OHCs and IHCs in the
2 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
presence of Tmc1-mCherry or Tmc2-AcGFP (Figure 1C). When
both transgenes were present, a small but statistically signiﬁcant
(p = 0.03) difference in MET currents was observed between
IHCs of wild-type and the Tmc1+2 BAC groups, but no signiﬁcant
difference (p = 0.50) was seen between OHCs of the two
groups. At P7, the effects of transgene expression differed between
IHCs and OHCs: in OHCs, the presence of Tmc1-mCherry
restored MET currents, but the presence of Tmc2-AcGFP did
not. The partial rescue of MET currents by Tmc2-AcGFP at P3,
but not at P7, is consistent with the previously reported decline
in Tmc2 mRNA levels in cochlear hair cells after early postnatal
increase (Kawashima et al., 2011), indicating an agreement between
spatiotemporal expression patterns of the transgenes in
rescued mice and the endogenous genes in wild-type mice.
The data also support the hypothesis that, when it is expressed,
TMC2 is able to partially compensate for TMC1 function (Kawashima
et al., 2011). In IHCs at P7, however, the presence of either
Tmc1-mCherry or Tmc2-AcGFP resulted in partial recovery of
MET currents, and expression of both transgenes resulted in a
complete recovery (Figure 1C). This suggests that Tmc2 expression
does not decline in IHC to the extent that it does in OHC
between P3 and P7. Collectively, these results show that
the Tmc1-mCherry and Tmc2-AcGFP transgenes can restore
MET function and normal ABR thresholds in Tmc1D/D
;Tmc2D/D
Figure 1. Rescue of Hair Cell Function in
Tmc1D/D
;Tmc2D/D
Mice Expressing TMC1mCherry
and TMC2-AcGFP
(A) Bacterial artiﬁcial chromosome (BAC) transgenes.
The stop codons of Tmc1 or Tmc2 were
replaced with cDNA encoding mCherry or AcGFP,
respectively.
(B) Mean (±SD) ABR thresholds of 6- to 8-week-old
mice for tone-burst stimuli of 8, 16, or 32 kHz. The
hearing in mice lacking functional TMC1 was
rescued by Tmc1-mCherry fully (line 1, 2, and 4) or
partially (line 3 females), but not by Tmc2-AcGFP.
(C) Mean (±SEM) maximal mechanotransduction
currents recorded from hair cells excised from
mice of the indicated genotypes. Mice with transgenes
are on Tmc1D/D
;Tmc2D/D
background.
Transduction currents were evoked using saturating
hair bundle deﬂections and were recorded
at À84 mV using the whole-cell, tight-seal technique.
Number of hair cells is indicated for each
genotype excised from three (wild-type), one
(Tmc1D/D
;Tmc2D/D
), three (Tmc1-mCherry), ﬁve
(Tmc2-AcGFP), and two (Tmc1-mCherry:Tmc2AcGFP)
P3 mice and three (wild-type), two
(Tmc1D/D
;Tmc2D/D
), four (Tmc1-mCherry), ﬁve
(Tmc2-AcGFP), and one (Tmc1-mCherry:Tmc2AcGFP)
P7 mice.
mice and that different transgene copy
numbers appear to have no signiﬁcant effect
on MET function or ABR thresholds.
Localization of TMC1
In Tmc1-mCherry mice on a Tmc1D/D
background, we observed mCherry ﬂuorescence
in the form of diffraction-limited puncta in stereocilia
of auditory and vestibular hair cells (Figures 2A–2D). The presence
of discrete puncta, rather than a uniform diffuse ﬂuorescence
along the stereocilia, indicates an enrichment of a few
closely associated TMC1 molecules in small clusters within
each punctum. In cochlear hair cells at P8 (Figures 2A and 2B),
TMC1-mCherry puncta were observed at the tips of the shorter
rows of stereocilia but were absent or rarely detected at the
tips of the tallest stereocilia. A few puncta were observed along
the length of stereocilia, suggesting that there may be a population
of TMC molecules not located at the site of MET. We also
noted that the TMC1-mCherry ﬂuorescence signal was absent
from the tips of a few stereocilia in the shorter rows (Figure 2A,
arrows). This may be due to decay or bleaching of the ﬂuorophores
or local damage of the stereocilia. Another possibility is
that some MET sites lack TMC1, at least transiently, and that
the presence of TMC is not required to sustain stereocilia
morphology. The pattern of TMC1-mCherry puncta appeared
to be indistinguishable between TMC1-mCherry mice on either
Tmc1D/D
(Figure 2A) or Tmc1D/+
backgrounds (Figure S1A), suggesting
that the presence of endogenous protein does not affect
or preclude the localization of mCherry-tagged TMC1 at stereocilia
tips. This pattern of predominant distribution of TMC1mCherry
at the tips of IHC and OHC stereocilia did not change
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 3
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
signiﬁcantly after exposure to 5 mM BAPTA for 15 min (Figure
S2), which is known to disrupt the tip links and abolish
MET (Assad et al., 1991). Similar to the observations in cochlear
hair cells, TMC1-mCherry puncta localized predominantly at the
tips of vestibular hair cell stereocilia (Figures 2C and 2D).
To conﬁrm the TMC1-mCherry localization results, we generated
antibodies to detect endogenous protein. The speciﬁcity of
Figure 2. Localization of TMC1-mCherry in
Hair Cell Stereocilia
(A) Confocal ﬂuorescence image of IHCs from
Tmc1-mCherry transgenic mice at P8 (line 4).
TMC1-mCherry (red) localizes predominantly to
the tips of the shorter rows of stereocilia, counterstained
in green with Alexa 488-phalloidin.
(B) TMC1-mCherry (red) distribution in mouse
OHC stereocilia (green) at P8 (line 2).
(CandD)TMC1-mCherry(red)localizestothetipsof
mouse vestibular stereocilia (green) at P7 (line 2).
(E) Immunoﬂuorescence confocal image of antiTMC1
antibody (PB277) in P9 mouse OHC
stereocilia.
(F) Close-up view of OHC stereocilia labeled with
PB277 shows that TMC1 labeling (red) is present
at the tips of second and third rows of stereocilia.
Lining up of TMC1 immunoﬂuorescence puncta at
the tips of stereocilia is clearly visualized in the
boxed region in (F) and magniﬁed below.
(G and H) Immunoﬂuorescence confocal images
of anti-TMC1 antibody (PB277) in P15 rat IHC (G)
and OHC (H) stereocilia also shows TMC1 labeling
(red) at the tips of second and third rows.
(I) TMC1 (red) is consistently detected at the tips of
P15 rat vestibular hair cell stereocilia (green) using
antibody PB612.
The scale bars represent 2 mm. See also Figures
S1–S3.
the antibodies for TMC1 was veriﬁed in
COS7 cells transfected with constructs
encoding TMC1 and TMC2 and in stereocilia
from Tmc1D/D
mice (Figure S3). To
detect and localize native TMC1 in
cochlear and vestibular stereocilia, we
used immunostaining based on a protocol
(Grati and Kachar, 2011) optimized
to detect immunoﬂuorescence signal
from low-copy-number proteins within
the crowded molecular environment of
the tip-link insertion site (Kachar et al.,
2000). We labeled both mouse and rat
inner ear tissue to verify the localization
of endogenous TMC proteins in an
additional species. Consistent with the
localization of TMC1-mCherry, TMC1
immunoﬂuorescence was observed at
the tips of the second and third row stereocilia
of mouse OHCs, but not at the
tips of the tallest row (Figures 2E and
2F). The same pattern of TMC1 immunoﬂuorescence
was also observed in rat
IHCs (Figure 2G) and OHCs (Figure 2H). Once again, we
observed that the immunoﬂuorescence signal was absent in
some shorter stereocilia (Figures 2E–2H). Whereas this may be
because TMC1 can be absent in some stereocilia tips, we cannot
exclude that their absence may have been due to epitope masking
when labeling a few protein molecules within the densely
packed (Kachar et al., 2000) lower-tip-link insertion site. TMC1
4 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
immunoﬂuorescence puncta were also observed predominantly
at the tips of vestibular hair cell stereocilia (Figure 2I).
Distinction between mCherry Fluorescence and
Autoﬂuorescence
It is noteworthy that some autoﬂuorescence was detected in
wild-type bundles not expressing TMC1-mCherry (Figure S1B).
This is likely due to the high gain setting of the system and
long exposure times needed to detect the low-abundance
TMC1-mCherry. Furthermore, the presence of endogenous
autoﬂuorescent molecules is exacerbated in stressed or dying
tissue (Buschke et al., 2012). To reduce bleaching of TMC1mCherry
and TMC2-AcGFP and limit autoﬂuorescence, we
performed rapid dissection and ﬁxation. The residual autoﬂuorescence
signals, which were clearly much lower in number
compared to TMC1-mCherry signal (Figure S1A), were likely
the result of debris deposited on the stereocilia surface during
dissection of the tissue. The distinction between TMC1-mCherry
signal and background autoﬂuorescence was clearly evident
when we examined organ of Corti tissue from Tmc1-mCherry
line 3 mice. Consistent with X-linked inactivation of the Tmc1mCherry
transgene, we observed a mosaic expression of
TMC1-mCherry ﬂuorescence in OHCs (Figures 3A and 3B) and
IHCs (Figures 3C and 3D) of female mice from this line. A similar
expression pattern was previously described for a Myo7a transgene
inserted in the X chromosome (Prosser et al., 2008). Quantifying
the number of ﬂuorescence puncta within hair bundles of
Tmc1-mCherry line 3 mice, we found that some bundles showed
puncta predominantly at the tips of stereocilia of the second and
shorter rows (Figures 3A and 3B). The average number of puncta
Figure 3. Mosaic Expression of TMC1-mCherry in Mouse Line 3
(A) Confocal image of OHC stereocilia (green) from P10 mice, showing mosaic expression of TMC1-mCherry ﬂuorescent puncta (red). Cells that are not
expressing TMC1-mCherry are highlighted with asterisks.
(B) Corresponding red channel (TMC1-mCherry) shown in gray.
(C) Stereocilia (green) from IHCs, of which only one cell (asterisk) is expressing TMC1-mCherry (red).
(D) Corresponding red channel (TMC1-mCherry) shown in gray.
(E–G) Confocal images of P10 mouse cochlear hair cells exposed to 5 mM FM1-43 FX. IHCs and OHCs from male Tmc1-mCherry line 3 mouse
(E), from female Tmc1-mCherry line 3 mouse (F), and from Tmc1D/D
;Tmc2D/D
mouse (G). The Tmc1-mCherry line 3 mice were on a Tmc1D/D
;Tmc2D/D
background.
The scale bars represent (A–D) 5 mm and (E–G) 100 mm. See also Figure S1.
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 5
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
per bundle was 44.8 ± 6 (n = 6 cells). The other bundles (Figures
3A and 3B, asterisks) showed a lower number (7.2 ± 3.5) of
randomly distributed puncta per bundle (n = 5 cells), which we
assume are due to background ﬂuorescence. Matching this
mosaic pattern of expression of TMC1-mCherry, a mosaic
pattern of FM1-43 dye uptake was also observed in P10 cochlear
hair cells from line 3 female mice (Figures 3E–3G). This was
consistent with the large variability in ABR thresholds measured
among these mice (Figure 1B).
Localization of TMC2
Similar to TMC1-mCherry, TMC2-AcGFP expressed in Tmc2AcGFP
transgenic mice on a Tmc2D/D
background was also
observed in stereocilia as diffraction-limited puncta (Figures
4A–4E). Consistent with the reported spatiotemporal expression
of Tmc2 mRNA (Kawashima et al., 2011), the expression of
TMC2-AcGFP in auditory hair cells is transient and disappears
as the hair cells mature. To assess the temporal expression during
organ of Corti development, we focused our analyses to the
mid-apical turn of the cochlea, i.e., 30%–50% from the apex of
the cochlea. In IHCs at P4, TMC2-AcGFP ﬂuorescence puncta
were localized to the tips of the second and shorter rows of sterFigure
4. Localization of TMC2-AcGFP in
Hair Cell Stereocilia
(A) Confocal images of IHC stereocilia (labeled in
red with Alexa Fluor 568 phalloidin) from Tmc2AcGFP
transgenic mice at P4 (line 4), showing
localization of TMC2-AcGFP (green).
(B and C) TMC2-AcGFP (green) localizes at the tips
of vestibular hair cell stereocilia (red) at P7 (line 2).
(D) In addition to stereocilia tip labeling, some hair
bundles at P7 also show abundant TMC2-AcGFP
puncta along the length of stereocilia.
(E) The expression of TMC2-AcGFP (green) within
stereocilia (red) was seen to vary from cell to cell in
vestibular organs (line 4; P8).
(F and G) Immunoﬂuorescence confocal images of
rat IHC and OHC stereocilia (green) at P7, showing
immunoﬂuorescence of TMC2 (red) using antiTMC2
antibody PB361.
(H) Immunoﬂuorescence staining of TMC2 (red) in
stereocilia (green) of rat vestibular hair cells at P3.
The scale bars represent (A–D and F–H) 2 mm and
(E) 5 mm. See also Figure S3.
eocilia and absent from the tips of the tallest
row (Figure 4A). Some puncta were
also randomly distributed along the stereocilia,
distant from the site of MET (Figure
4A). Similar to TMC1-mCherry, we
also noted that the TMC2-AcGFP ﬂuorescence
signal was absent from the tips of
some stereocilia in the shorter rows (Figure
4A, arrows). Localization of TMC2AcGFP
puncta at the tips of stereocilia
was also observed in vestibular hair cells
(Figures 4B and 4C). However, when
compared to cochlear hair cells, greater
variations in both the expression level
and pattern of distribution of TMC2-AcGFP puncta were
observed between stereocilia bundles of different vestibular
hair cells, with much-larger numbers of puncta present along
the length of stereocilia in some cells (Figures 4D and 4E). These
differences may be correlated with variations in the type, developmental
stage, or both of the different hair cells in the vestibular
sensory epithelia or may underlie previously unrecognized functional
variations between them. Alternatively, they may be the
result of transgene overexpression. Immunoﬂuorescence using
antibodies speciﬁc for TMC2 (Figure S3) in rat inner ear tissue
conﬁrmed localization of TMC2 at the tips of stereocilia in
IHCs, OHCs, and vestibular hair cells (Figures 4F–4H). The
observation of stereocilia tip localization of TMC2 in rat cochlear
hair cells as late as P15 (Figures 4F and 4G) may be due to differences
in the time course of Tmc2 expression in rat hair cells
compared with mouse hair cells.
Spatial and Temporal Colocalization of TMC1 and TMC2
We next investigated the temporal expression and extent of
colocalization of TMC1-mCherry and TMC2-AcGFP during postnatal
development in stereocilia of mice expressing both transgenes
on a null Tmc1D/D
;Tmc2D/D
background (Figures 5 and
6 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
S4–S6). The earliest that we could consistently detect TMC1mCherry
or TMC2-AcGFP in auditory hair cell stereocilia was
at P2 (Figure 5A), when TMC2-AcGFP signal was quite evident.
By P3, both TMC1-mCherry and TMC2-AcGFP ﬂuorescent
puncta were abundant and localized along the length of stereocilia
as well as at the tips, sometimes overlapping (Figure 5A). At
P7, TMC2-AcGFP puncta were less evident in OHC stereocilia
(Figure S4), but both TMC1-mCherry and TMC2-AcGFP puncta
were equally present at stereocilia tips of IHCs (Figure 5A). At
P10, TMC2-AcGFP puncta also disappeared from IHC stereocilia,
whereas TMC1-mCherry puncta persisted at stereocilia tips
of both IHCs (Figure 5A) and OHCs (Figure S4).
To examine the extent of overlap of localization of TMC1 and
TMC2, we analyzed confocal images of IHCs at P7, where
both TMC1-mCherry and TMC2-AcGFP were equally abundant
(Figures 5A and 5B). The TMC1-mCherry and TMC2-AcGFP
puncta were ﬁrst identiﬁed as local, high-intensity regions (maxima)
in the image, using a custom written code in MATLAB. At
each of the identiﬁed local maxima, the center of mass of image
intensity under a disc-shaped mask roughly the size of the point
spread function (ﬁve-pixel radius) was then computed to obtain
the centroids of each punctum (Figure 5B). The distance between
the computed centroid of each TMC1-mCherry punctum
and its closest TMC2-AcGFP punctum was used as a measure
of separation between puncta. A histogram of the frequency distribution
of puncta as a function of their distance to their closest
neighbor is shown in Figure 5C. Approximately 25% of the ﬂuorescent
TMC1-mCherry puncta had a TMC2-AcGFP punctum
that was within a 30-nm radius and thus potentially part of the
same molecular aggregate or complex (number of puncta =
709; number of cells = 12). Similar results were observed when
plotting the frequency of TMC2-AcGFP puncta as a function of
the distance to their closest TMC1-mCherry puncta (Figure S5A).
Many of the overlapping TMC1-mCherry and TMC2-AcGFP
Figure 5. Simultaneous Localization of
TMC1-mCherry and TMC2-AcGFP in Mouse
Cochlear and Vestibular Stereocilia
(A) Confocal images of IHC stereocilia from
transgenic mice expressing TMC1-mCherry and
TMC2-AcGFP at P1, P2, P3, P7, and P10 (Tmc1mCherry
line 2; Tmc2-AcGFP line 3). All cochlear
stereocilia bundles are from the mid-apical region
of the cochlea (located between 30% and 50%
from the apex of the cochlea). Stereocilia (labeled
with Alexa Fluor 405 phalloidin) are shown in white
(left). Right panels show, respectively, merge of
red and green channels, only red channel, and only
green channel.
(B) IHC stereocilia at P7, with calculated centroids
for TMC1-mCherry and TMC2-AcGFP puncta
represented as red and green dots, respectively,
to show relative localization of both signals.
(C) Histogram showing the frequency distribution
of TMC1-mCherry puncta centroids according
to their distance from the centroid of the corresponding
closest neighboring TMC2-AcGFP
punctum (histogram bin size = 30 nm; number of
puncta = 709; number of hair cells = 12).
(D) Scatterplot of ﬂuorescence intensity (in a.u.) of
each TMC1-mCherry puncta as a function of the
distance from their corresponding closest neighboring
TMC2-AcGFP punctum.
(E) Confocal images of vestibular hair cells from
P13 mice expressing TMC1-mCherry and TMC2AcGFP.
Stereocilia, labeled with Alexa-Fluor-405conjugated
phalloidin, are shown in white (Tmc1mCherry
line 2; Tmc2-AcGFP line 4).
(F) Close-up views of vestibular stereocilia bundles
from P3 mouse expressing both TMC1-mCherry
and TMC2-AcGFP showing presence of only
TMC1-mCherry (red arrows), only TMC2-AcGFP
(green arrows), or both (yellow arrows) at stereocilia
tips (Tmc1-mCherry line 2; Tmc2-AcGFP
line 4).
The scale bars represent (A–D and F) 2 mm and (E)
5 mm. See also Figures S4–S6.
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 7
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
puncta localize to the tips of the second row and shorter stereocilia
(Figure 5B, arrows).
We next wanted to determine whether TMC1 and TMC2
compete with each other as interchangeable subunits of a heteromeric
aggregate. To do this, we ﬁrst calculated the intensities
of each punctum by integrating the pixel intensity within a point
spread function-sized disc (ﬁve-pixel radius; pixel size = 30 nm).
We then plotted the ﬂuorescence intensities of each TMC1mCherry
punctum as a function of the distance to its closest
TMC2-AcGFP neighbor (Figure 5D). The average intensity of
TMC1-mCherry puncta that had a TMC2-AcGFP punctum
neighbor within 30 nm was 5.86 ± 2.1, in a.u. (number of puncta =
93; number of cells = 10). The average intensity of TMC1mCherry
puncta whose nearest TMC2-AcGFP neighbor was
more than 100 nm away was 4.84 ± 1.7 (number of puncta =
331; number of cells = 10). Conversely, we also plotted the ﬂuorescence
intensities of the TMC2-AcGFP puncta as a function of
the distance of its closest TMC1-mCherry neighbor (Figure S5B).
The average intensity of TMC2-AcGFP puncta that had a TMC1mCherry
punctum neighbor within 30 nm was 8.1 ± 3.8 (number
of puncta = 93; number of cells = 10). The average intensity of
TMC2-AcGFP puncta when the distance to the closest TMC1mCherry
neighbor was more than 100 nm was 7.5 ± 3.9 (number
of puncta = 399; number of cells = 10). These data collectively
suggest that the amount of each protein was not reduced
when both proteins co-existed within a 30 nm range and potentially
as part of the same molecular cluster.
In the vestibular organs, we did not observe temporal expression
differences between TMC1-mCherry and TMC2-AcGFP
from P3 (earliest point we examined) to adulthood, and both
TMC1-mCherry and TMC2-AcGFP showed a higher degree of
cell-to-cell variability compared to cochlear hair cells. Whereas
most stereocilia bundles expressed both proteins to some
degree, some cells showed predominance of either TMC1mCherry
or TMC2-AcGFP (Figure 5E). When both were expressed
in the same hair cell, we observed partial overlap of
TMC1-mCherry and TMC2 ﬂuorescence at stereocilia tips (Figure
5F). Of note, the ﬂuorescence puncta sometimes appear to
be slightly separated above the stereocilia tip. This apparent
mismatch is the combined result of the differential size of the
point spread function for the ﬂuorophores and for the actin counterstained
stereocilia (stereocilia were stained with phalloidin
conjugated to Alexa Fluor 405, whose emission wavelength is
much shorter than the emission wavelengths of mCherry or
AcGFP nm) with the fact that the stereocilia often have an elongated
and thinner-beveled tip where the membrane can be
slightly pulled away or tented by tension on the tip link (see Figure
4E in Kachar et al., 2000). Together, these parameters give
the false impression that the TMC ﬂuorescence puncta are separated
from the tip of the actin-counterstained stereocilia.
TMC proteins have been reported to interact with PCDH15
isoforms, and PCDH15 also interacts with CDH23 to form the kinociliary
links between the tallest stereocilia and the kinocilium
(Gillespie et al., 2005; Goodyear et al., 2010; Kazmierczak
et al., 2007). Based on this, we looked for localization of
TMC1-mCherry and TMC2-AcGFP along the tallest stereocilia
in the bundle but did not detect any enrichment of either protein
at this region.
Whereas the stereocilia tip localization of TMC1 and TMC2
support their having a role in mechanotransduction, we next
investigated how TMC1-mCherry and TMC2-AcGFP ﬂuorescence
intensity compares between hair cell body and stereocilia
to explore the possibility for additional roles of these proteins
elsewhere in the cell. The level of ﬂuorescence signal from the
cell bodies of hair cells expressing TMC1-mCherry and TMC2AcGFP
on a Tmc1D/D
;Tmc2D/D
background (Figures S6A and
S6B, asterisk) was difﬁcult to differentiate from autoﬂuorescence,
which is also present in neighboring supporting cells
that are not believed to express the tagged proteins (Figures
S6A and S6B, double asterisks). When we compared ﬂuorescence
in the red channel between hair cells from Tmc1-mCherry
mice and wild-type mice (Figures S6C–S6F), a distinct ﬂuorescence
signal was detected in the hair cell bodies (Figures
S6C–S6F, asterisk) of the transgenics compared to wild-type
controls under identical image acquisition conditions. This ﬂuorescence
likely corresponds to TMC1-mCherry in endocytic or
ER-associated vesicles. No enhancement of ﬂuorescence signal
was detected in the hair cell basolateral membrane.
Reverse-Polarity Currents in Tmc1D/D
;Tmc2D/D
Mice
Recently, non-canonical, reverse-polarity MET currents evoked
by deﬂecting the hair bundle in the direction away from the tallest
stereocilia (inhibitory) were recorded from hair cells (Beurg et al.,
2014; Kim et al., 2013; Marcotti et al., 2014). These reverse-polarity
currents have been shown to occur after tip-link disruption
by the calcium-chelating agent BAPTA (Kim et al., 2013; Marcotti
et al., 2014). Whereas their physiological signiﬁcance and underlying
mechanisms are yet to be elucidated, it was suggested that
these reverse-polarity currents may be caused by the relocation
of the MET channel as a result of tip link disruption, which might
allow for its activation by inhibitory stimuli (Barr-Gillespie and
Nicolson, 2013). Because the TMC proteins are hypothesized
to be associated with MET, we investigated whether these proteins
are required for the reverse-polarity MET currents (Beurg
et al., 2014; Kim et al., 2013; Marcotti et al., 2014).
We ﬁrst conﬁrmed the presence of the reverse-polarity MET
currents in wild-type hair cells following treatment with 5 mM
BAPTA (Figure 6A). We next examined our Tmc1D/D
;Tmc2D/D
double mutants that are veriﬁed null alleles (Kawashima et al.,
2011) that do not exhibit conventional transduction when stimulated
with a glass probe (Figure 1). We did not detect any conventional
transduction upon delivery of a 40-Hz sine wave
stimulus using ﬂuid jet simulation. However, a 25%–50% increase
in the sine wave stimulus amplitude, which in wild-type
hair cells evoked maximum MET currents, resulted in appearance
of the reverse-polarity currents (Figure 6B). Our data are
consistent with recent reports that recorded reverse-polarity
currents in Tmc1 and Tmc2 mutant mice (Beurg et al., 2014;
Kim et al., 2013).
DISCUSSION
Whereas earlier studies suggested that TMC1 is involved in molecular
trafﬁcking or intracellular regulatory signaling for hair cell
differentiation (Marcotti et al., 2006), more-recent studies have
provided genetic and physiological evidence for a direct role
8 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
for TMC1 and TMC2 in hair cell MET (Kawashima et al., 2011;
Kim et al., 2013; Pan et al., 2013), suggesting that these proteins
are components of the MET complex. One of these reports (Pan
et al., 2013) further suggests that TMC1 and TMC2 may be components
of the pore-forming element, and the varying conductance
properties of the channel may be the result of variations
in the relative stoichiometry of these proteins in the channel complex.
However, an important milestone in corroborating this hypothesis
(Liedtke, 2014) is verifying the precise localization of
these proteins at the site of MET (Beurg et al., 2009). The current
study shows that TMC1 and TMC2 localize to the speciﬁc site of
MET at stereocilia tips. We addressed this outstanding question
by using two different approaches to detect the proteins. First,
we generated transgenic mice expressing mCherry-tagged
TMC1 and AcGFP-tagged TMC2. Second, we generated antibodies
speciﬁc for native TMC1 and TMC2 that we used to label
mouse and rat inner ear tissue.
Mice expressing both TMC1-mCherry and TMC2-AcGFP on a
Tmc1D/D
;Tmc2D/D
background showed normal vestibular phenotypes,
normal hearing thresholds, and normal whole-cell
MET currents within the ranges reported for mice segregating
different numbers of wild-type alleles of endogenous Tmc1
and Tmc2 (Pan et al., 2013). These observations demonstrate
that the ﬂuorophore-tagged proteins are functional, properly
temporally expressed, and targeted to their site of function.
Our spatiotemporal characterization of TMC1 and TMC2 localization
in cochlear hair cells provides insights into the intricate
and complex nature of MET currents. In the cochlea at P1,
neither TMC1-mCherry nor TMC2-AcGFP is detectable in the
stereocilia. By P3, both are abundantly expressed along the
length and at the tips of stereocilia. The overall expression of
both proteins then declines until about P7, with localization
now predominantly at the tips of stereocilia, and a much-smaller
population of TMC puncta still present along the stereocilia.
TMC2-AcGFP expression declines until it virtually disappears
by P10, whereas TMC1-mCherry remains at stereocilia tips
through to adulthood. The observed spatiotemporal reﬁnement
of TMC localization is reminiscent of the pruning of extracellular
stereociliary links and, in particular, lateral links and ankle links,
which are present in early postnatal hair bundles but progressively
disappear, leaving only tip links and top connectors in
mature hair cells (Goodyear et al., 2005; Waguespack et al.,
2007). Furthermore, the developmentally correlated changes in
TMC1-mCherry and TMC2-AcGFP localization are consistent
with the reported spatiotemporal expression of Tmc1 and
Tmc2 mRNA (Kawashima et al., 2011) as well as the acquisition
of MET in mouse cochlear hair cells (Lelli et al., 2009). In cochlear
hair cells, the TMC proteins were rarely detected at the tips of the
tallest row of stereocilia, where functional MET channels are
thought to be absent. In vestibular organs, both TMC1-mCherry
and TMC2-AcGFP localize to stereocilia tips of mature hair cells
showing a cell-to-cell variability in the degree of colocalization of
these proteins. Some hair cells show more TMC1 whereas
others show more TMC2. The localization pattern of the ﬂuorophore-tagged
TMC1 and TMC2 was conﬁrmed in stereocilia
from mice and rats by immunoﬂuorescence detection of endogenous
proteins using antibodies speciﬁc to TMC1 or TMC2.
We observed a variable number of TMC1-mCherry and TMC2AcGFP
puncta along the length of all mature hair cell stereocilia,
which may be due to increased copy number of the proteins in
the transgenic mice. However, another possibility is that hair
cells retain a pool of TMC proteins along the stereocilia plasma
membrane that are actively trafﬁcking or have a yet-to-be-determined
function. That we were able to measure reverse-polarity
MET currents recently reported by several groups (Barr-Gillespie
and Nicolson, 2013; Kim et al., 2013; Marcotti et al., 2014) in our
double homozygous null mice (Tmc1D/D
;Tmc2D/D
) conﬁrms that
neither TMC1 nor TMC2 is required for these currents. It is worth
noting that we do not detect enrichment of TMC1 or TMC2 along
the taller stereocilia contacting the kinocilium where PCDH15 is
concentrated, in either developing cochlear hair cells or in
vestibular hair cells (Kazmierczak et al., 2007). This suggests
that interaction of TMC1/2 with PCDH15 is limited to the site of
MET at the tips of stereocilia and does not involve kinociliary
Figure 6. Reverse-Polarity Currents in WT
and Tmc1D/D
;Tmc2D/D
Hair Cells
(A) Example of MET current (black trace) recorded
from a WT OHC (from mid-apical turn of the
cochlea; P7 mouse) in response to a sine wave
stimulus (black) with a ﬂuid jet device. Upon
application of 5 mM BAPTA, the MET current of the
same cell is abolished (red trace) and followed by
the progressive appearance of the reverse-polarity
currents.
(B) Representative MET current recordings from
a Tmc1D/D
;Tmc2D/D
OHC. No conventional MET
currents were detected during regular stimulation
(black trace). A higher stimulus (red sine wave)
evoked reverse-polarity currents (red trace). All
recordings in (A) and (B) are averages of three to
eight individual traces (n = 5 cells).
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 9
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
links. Inconsistent with our data, a recent attempt to localize
TMC1 by immunoﬂuorescence using an antigen retrieval method
in wild-type mice (Beurg et al., 2015) showed immunoreactivity
along the length of the stereocilia, at the tips of the tallest rows
of stereocilia, and in the kinocilium—sites where no mechanotransduction
is thought to occur.
There are currently no direct data on molecular-level interactions
of TMC1 and TMC2. Distinct putative single-channel conductances
recently reported from wild-type IHCs (Pan et al.,
2013) led the authors to propose that variations in the stoichiometry
of TMC1 and TMC2 might explain tonotopic gradients in the
conductance properties of the MET channel. Our localization
study provides evidence that TMC2 is only transiently present
in stereocilia of cochlear hair cells and thus might not contribute
to tonotopic gradients in channel properties (Pan et al., 2013) in
mature hair bundles. However, different conductance levels can
arise from differential abilities of TMC1 and TMC2, alone or
together, to interact with other proteins that affect single-channel
MET conductance (Beurg et al., 2014), simultaneous gating of
multiple MET channels in the same complex on a single stereocilium,
or different channel complexes on different stereocilia.
Figure 7. Schematic Diagram of Stereocilia
MET Channel Complex Proteins
Diagram of the stereocilia MET channel complex
illustrating the localization of known associated
proteins. Localization of TMC1 and TMC2 is
consistent with these proteins being components
of the complex.
Our results also show that, when both
proteins are expressed at about the
same levels, as in IHC stereocilia at P7,
only $25% of TMC1-mCherry puncta
have a TMC2-AcGFP punctum located
within a 30-nm radius and therefore
potentially within the same molecular
complex or cluster. The remaining puncta
were randomly dispersed along the stereocilia.
It is possible that TMC1 and
TMC2 proteins trafﬁc to the stereocilia
membrane as independent clusters or
homo-oligomers and co-accumulate in
the limited space at the stereocilia tips
by interacting with other MET components
such as PCDH15 (Maeda et al.,
2014). This interpretation is supported
by our observation that the ﬂuorescence
intensity signal emitted by either TMC1mCherry
or TMC2-AcGFP was on
average the same regardless of whether
the puncta were within interacting distance
(<30 nm) or not (>100 nm). This
observation may indicate that the presence
of one TMC does not exclude the
presence of the other at the same locus.
It may also indicate that the quantities of
TMC1 and TMC2 do not vary whether or
not the other is present. One could interpret this to argue against
a role for these proteins as pore-forming elements of the MET
channel (Pan et al., 2013).
In our current study localizing TMC1 and TMC2 to the site of
canonical MET in hair cell stereocilia strongly supports the hypothesis
that they are members of the MET channel complex
(schematic model in Figure 7). Whether they are in fact the
MET channel or non-pore-forming subunits/accessory proteins
modulating the kinetics of gating, channel properties, stability,
or subcellular targeting of the actual channel is yet to be conclusively
established. Association of TMC1 and TMC2 with other
MET proteins to form a stable complex may explain TMC1
enrichment at stereocilia tips even after tip link disruption by
BAPTA. Similar to TMC, PCDH15 remains at stereocilia tips
following BAPTA treatment (Indzhykulian et al., 2013). Other
proteins, such as PCDH15 and TMIE, which were recently
shown to interact with TMC proteins (Beurg et al., 2015; Fuchs,
2015; Maeda et al., 2014), and TMHS/LHFPL5 identiﬁed as
an MET channel accessory component (Xiong et al., 2012)
may contribute to the spatiotemporal distribution of TMC1 and
TMC2 in stereocilia.
10 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
EXPERIMENTAL PROCEDURES
Generation of Transgenic Mice
We used bacterial artiﬁcial chromosomes (BACs) encoding mouse Tmc1 or
Tmc2 genes. Details of the BACs and procedures for their modiﬁcation,
puriﬁcation, and injection into mouse eggs to generate the transgenic mice,
as well as backcrossing onto mice segregating targeted deletion alleles of
Tmc1 and Tmc2, are fully described in the Supplemental Experimental Procedures.
Founder mice and offspring carrying the BAC transgenes were identiﬁed
by PCR analysis of genomic DNA isolated from tail biopsies. All animal
protocols and procedures were approved by a NIH Animal Care and Use
Committee.
Hair Cell Electrophysiology
Whole-cell, tight-seal technique was used to record mechanotransduction
currents from P3–P8 mice by deﬂecting the hair bundles of IHCs and OHCs using
a stiff glass probe. To measure reverse-polarity currents, hair bundles were
mechanically stimulated with a ﬂuid jet system. The position of the stimulating
pipette was adjusted to elicit a maximal MET current. All experiments were
performed at room temperature. Details are provided in Supplemental Experimental
Procedures.
Detection of TMC1-mCherry and TMC2-AcGFP and
Immunoﬂuorescence Microscopy
Inner ear tissue from mice expressing TMC1-mCherry and/or TMC2-AcGFP
was dissected and mounted between slide and cover slip and viewed in a
Nikon inverted ﬂuorescence microscope, outﬁtted with a spinning disk
confocal scan head, 1003 Apo TIRF 1.49 N.A. objective, and an EM-CCD
camera. NIS-Elements imaging software was utilized for image acquisition
and analysis.
For immunoﬂuorescence, wild-type mice and rat tissue was labeled using
polyclonal antibodies against mouse TMC1 (PB277 and PB612) and mouse
TMC2 (PB361). See Supplemental Information for further details.
SUPPLEMENTAL INFORMATION
Supplemental Information includes Supplemental Experimental Procedures,
six ﬁgures, and four tables and can be found with this article online at http://
dx.doi.org/10.1016/j.celrep.2015.07.058.
AUTHOR CONTRIBUTIONS
K.K. conceived, designed, and generated transgenic mice and collected and
analyzed data. S.E. collected and analyzed data. B.P. collected and analyzed
conventional MET electrophysiology data. M.S. collected and analyzed electrophysiology
data acquired with ﬂuid jet stimulation. B.A.M. helped acquire
confocal microscope images. P.S. helped with analyses and quantiﬁcation
of ﬂuorescence data. R.C., H.N., T.F., and Y.K. helped collect data. B.Y.C.
and K.M. helped generate transgenic mice. J.R.H. analyzed conventional
MET data. A.J.G. conceived the overall study and designed transgenic mice.
S.E. and B.K. conceived the overall study and localization experiments,
collected and analyzed data, and generated the ﬁgures. K.K., S.E., A.J.G.,
and B.K. wrote the paper. All authors provided comments and approved the
ﬁnal submission of the manuscript.
CONFLICTS OF INTEREST
K.K. and A.J.G. hold the following U.S. patents: 7,166,433 (Transductin-2 and
Applications to Hereditary Deafness), 7,192,705 (Transductin-1 and Applications
to Hereditary Deafness), and 7,659,115 (Nucleic Acid Encoding Human
Transductin-1 Polypeptide).
ACKNOWLEDGMENTS
We thank Elizabeth Wilson, James McGehee, and Patrick Diers for mouse
management. Work was supported by NIH intramural research funds Z01DC000002
(to B.K.) and Z01-DC000060 (to A.J.G.). B.P. and J.R.H. were supported
by NIH/NIDCD RO1-DC013521.
Received: December 23, 2014
Revised: June 11, 2015
Accepted: July 28, 2015
Published: August 27, 2015
REFERENCES
Assad, J.A., Shepherd, G.M., and Corey, D.P. (1991). Tip-link integrity and
mechanical transduction in vertebrate hair cells. Neuron 7, 985–994.
Barr-Gillespie, P.G., and Nicolson, T. (2013). Who needs tip links? Backwards
transduction by hair cells. J. Gen. Physiol. 142, 481–486.
Beurg, M., Fettiplace, R., Nam, J.H., and Ricci, A.J. (2009). Localization of inner
hair cell mechanotransducer channels using high-speed calcium imaging.
Nat. Neurosci. 12, 553–558.
Beurg, M., Kim, K.X., and Fettiplace, R. (2014). Conductance and block of haircell
mechanotransducer channels in transmembrane channel-like protein mutants.
J. Gen. Physiol. 144, 55–69.
Beurg, M., Xiong, W., Zhao, B., Mu¨ ller, U., and Fettiplace, R. (2015). Subunit
determination of the conductance of hair-cell mechanotransducer channels.
Proc. Natl. Acad. Sci. USA 112, 1589–1594.
Buschke, D.G., Squirrell, J.M., Fong, J.J., Eliceiri, K.W., and Ogle, B.M. (2012).
Cell death, non-invasively assessed by intrinsic ﬂuorescence intensity of
NADH, is a predictive indicator of functional differentiation of embryonic
stem cells. Biol. Cell 104, 352–364.
Fuchs, P.A. (2015). How many proteins does it take to gate hair cell mechanotransduction?
Proc. Natl. Acad. Sci. USA 112, 1254–1255.
Gale, J.E., Marcotti, W., Kennedy, H.J., Kros, C.J., and Richardson, G.P.
(2001). FM1-43 dye behaves as a permeant blocker of the hair-cell mechanotransducer
channel. J. Neurosci. 21, 7013–7025.
Gillespie, P.G., Dumont, R.A., and Kachar, B. (2005). Have we found the tip
link, transduction channel, and gating spring of the hair cell? Curr. Opin. Neurobiol.
15, 389–396.
Goodyear, R.J., Marcotti, W., Kros, C.J., and Richardson, G.P. (2005). Development
and properties of stereociliary link types in hair cells of the mouse
cochlea. J. Comp. Neurol. 485, 75–85.
Goodyear, R.J., Forge, A., Legan, P.K., and Richardson, G.P. (2010). Asymmetric
distribution of cadherin 23 and protocadherin 15 in the kinocilial links
of avian sensory hair cells. J. Comp. Neurol. 518, 4288–4297.
Grati, M., and Kachar, B. (2011). Myosin VIIa and sans localization at stereocilia
upper tip-link density implicates these Usher syndrome proteins in mechanotransduction.
Proc. Natl. Acad. Sci. USA 108, 11476–11481.
Indzhykulian, A.A., Stepanyan, R., Nelina, A., Spinelli, K.J., Ahmed, Z.M., Belyantseva,
I.A., Friedman, T.B., Barr-Gillespie, P.G., and Frolenkov, G.I. (2013).
Molecular remodeling of tip links underlies mechanosensory regeneration in
auditory hair cells. PLoS Biol. 11, e1001583.
Kachar, B., Parakkal, M., Kurc, M., Zhao, Y., and Gillespie, P.G. (2000). Highresolution
structure of hair-cell tip links. Proc. Natl. Acad. Sci. USA 97, 13336–
13341.
Kawashima, Y., Ge´ le´ oc, G.S., Kurima, K., Labay, V., Lelli, A., Asai, Y., Makishima,
T., Wu, D.K., Della Santina, C.C., Holt, J.R., and Grifﬁth, A.J. (2011). Mechanotransduction
in mouse inner ear hair cells requires transmembrane
channel-like genes. J. Clin. Invest. 121, 4796–4809.
Kawashima, Y., Kurima, K., Pan, B., Grifﬁth, A.J., and Holt, J.R. (2015). Transmembrane
channel-like (TMC) genes are required for auditory and vestibular
mechanosensation. Pﬂugers Arch. 467, 85–94.
Kazmierczak, P., Sakaguchi, H., Tokita, J., Wilson-Kubalek, E.M., Milligan,
R.A., Mu¨ ller, U., and Kachar, B. (2007). Cadherin 23 and protocadherin 15
interact to form tip-link ﬁlaments in sensory hair cells. Nature 449, 87–91.
Kim, K.X., Beurg, M., Hackney, C.M., Furness, D.N., Mahendrasingam, S., and
Fettiplace, R. (2013). The role of transmembrane channel-like proteins in the
Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors 11
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
operation of hair cell mechanotransducer channels. J. Gen. Physiol. 142,
493–505.
Kurima, K., Peters, L.M., Yang, Y., Riazuddin, S., Ahmed, Z.M., Naz, S.,
Arnaud, D., Drury, S., Mo, J., Makishima, T., et al. (2002). Dominant and recessive
deafness caused by mutations of a novel gene, TMC1, required for
cochlear hair-cell function. Nat. Genet. 30, 277–284.
Lelli, A., Asai, Y., Forge, A., Holt, J.R., and Ge´ le´ oc, G.S. (2009). Tonotopic
gradient in the developmental acquisition of sensory transduction in outer
hair cells of the mouse cochlea. J. Neurophysiol. 101, 2961–2973.
Liedtke, W. (2014). A precisely deﬁned role for the tip link-associated protein
TMIE in the mechanoelectrical transduction channel complex of inner ear
hair cells. Neuron 84, 889–891.
Lyon, M.F. (1961). Gene action in the X-chromosome of the mouse (Mus musculus
L.). Nature 190, 372–373.
Maeda, R., Kindt, K.S., Mo, W., Morgan, C.P., Erickson, T., Zhao, H., ClemensGrisham,
R., Barr-Gillespie, P.G., and Nicolson, T. (2014). Tip-link protein protocadherin
15 interacts with transmembrane channel-like proteins TMC1 and
TMC2. Proc. Natl. Acad. Sci. USA 111, 12907–12912.
Marcotti, W., van Netten, S.M., and Kros, C.J. (2005). The aminoglycoside antibiotic
dihydrostreptomycin rapidly enters mouse outer hair cells through the
mechano-electrical transducer channels. J. Physiol. 567, 505–521.
Marcotti, W., Erven, A., Johnson, S.L., Steel, K.P., and Kros, C.J. (2006). Tmc1
is necessary for normal functional maturation and survival of inner and outer
hair cells in the mouse cochlea. J. Physiol. 574, 677–698.
Marcotti, W., Corns, L.F., Desmonds, T., Kirkwood, N.K., Richardson, G.P.,
and Kros, C.J. (2014). Transduction without tip links in cochlear hair cells is
mediated by ion channels with permeation properties distinct from those of
the mechano-electrical transducer channel. J. Neurosci. 34, 5505–5514.
Pan, B., Ge´ le´ oc, G.S., Asai, Y., Horwitz, G.C., Kurima, K., Ishikawa, K., Kawashima,
Y., Grifﬁth, A.J., and Holt, J.R. (2013). TMC1 and TMC2 are components
of the mechanotransduction channel in hair cells of the mammalian inner
ear. Neuron 79, 504–515.
Prosser, H.M., Rzadzinska, A.K., Steel, K.P., and Bradley, A. (2008). Mosaic
complementation demonstrates a regulatory role for myosin VIIa in actin dynamics
of stereocilia. Mol. Cell. Biol. 28, 1702–1712.
Sakaguchi, H., Tokita, J., Mu¨ ller, U., and Kachar, B. (2009). Tip links in hair
cells: molecular composition and role in hearing loss. Curr. Opin. Otolaryngol.
Head Neck Surg. 17, 388–393.
Schwander, M., Kachar, B., and Mu¨ ller, U. (2010). Review series: The cell
biology of hearing. J. Cell Biol. 190, 9–20.
Vreugde, S., Erven, A., Kros, C.J., Marcotti, W., Fuchs, H., Kurima, K., Wilcox,
E.R., Friedman, T.B., Grifﬁth, A.J., Balling, R., et al. (2002). Beethoven, a
mouse model for dominant, progressive hearing loss DFNA36. Nat. Genet.
30, 257–258.
Waguespack, J., Salles, F.T., Kachar, B., and Ricci, A.J. (2007). Stepwise
morphological and functional maturation of mechanotransduction in rat outer
hair cells. J. Neurosci. 27, 13890–13902.
Xiong, W., Grillet, N., Elledge, H.M., Wagner, T.F., Zhao, B.,Johnson,K.R.,Kazmierczak,
P., and Mu¨ ller, U. (2012). TMHS is an integral component of the
mechanotransduction machinery of cochlear hair cells. Cell 151, 1283–1295.
Zhao, B., Wu, Z., Grillet, N., Yan, L., Xiong, W., Harkins-Perry, S., and Mu¨ ller, U.
(2014). TMIE is an essential component of the mechanotransduction machinery
of cochlear hair cells. Neuron 84, 954–967.
12 Cell Reports 12, 1–12, September 8, 2015 ª2015 The Authors
Please cite this article in press as: Kurima et al., TMC1 and TMC2 Localize at the Site of Mechanotransduction in Mammalian Inner Ear Hair Cell Stereocilia,
Cell Reports (2015), http://dx.doi.org/10.1016/j.celrep.2015.07.058
Cell Reports
Supplemental Information
TMC1 and TMC2 Localize at the Site
of Mechanotransduction in Mammalian
Inner Ear Hair Cell Stereocilia
Kiyoto Kurima, Seham Ebrahim, Bifen Pan, Miloslav Sedlacek, Prabuddha Sengupta,
Bryan A. Millis, Runjia Cui, Hiroshi Nakanishi, Taro Fujikawa, Yoshi Kawashima, Byung
Y. Choi, Kelly Monahan, Jeffrey R. Holt, Andrew J. Griffith, and Bechara Kachar
Supplemental Data
Figure S1. Detection of TMC1-mCherry in Tmc1-mCherry mice on Tmc1Δ/+
background. (A) IHCs from Tmc1-mCherry transgenic mouse on Tmc1Δ/+
background
(Line 3, P10). TMC1-mCherry fluorescence puncta localized predominantly at stereocilia
tips. (B) Confocal image of IHC stereocilia (labeled with AlexaFluor488 phalloidin) from
a wild type mouse (P12), showing the presence of a few autofluorescence puncta. These
were weaker in intensity than TMC1-mCherry puncta, and were randomly distributed
(arrows). Related to Figures 2 and 3.
Figure S2. Stereocilia localization of TMC1-mCherry is seemingly unaffected by
treatment with BAPTA. Confocal fluorescence image of mid-apical OHCs (A) and
IHCs (B) from P10 transgenic mice expressing TMC1-mCherry (red) after 15 min
treatment with BAPTA, with actin counterstained with AlexaFluor488 (green). [Scale bar
= 5µm]. Related to Figure 2.
Figure S3. Validation of primary antibodies against TMC1 and TMC2. (A-C) COS7
cells transfected with Tmc1-mCherry and labeled with TMC1- (A and B) or TMC2-
specific (C) antibodies. (D-F) COS7 cells transfected with Tmc2-mCherry labeled with
TMC1- (D and E) or TMC2-specific (F) antibodies. (G-I) COS7 cells transfected with
Tmc4-mCherry labeled with TMC1- (G and H) or TMC2-specific (I) antibodies. (J-L)
COS7 cells transfected with Tmc5-mCherry labeled with TMC1- (J and K) or TMC2specific
(L) antibodies. Anti-TMC1 antibodies, PB277 and PB612, co-localized with
TMC1-mCherry, resulting in a yellow signal (A and B), but did not react with TMC2mCherry
(D and E), TMC4-mCherry (G and H) or TMC5-mCherry (J and K), resulting
in a red signal. Similarly, the anti-TMC2 antibody PB361 was specific for TMC2mCherry,
resulting in a yellow signal (F), but did not localize with TMC1-mCherry (C),
TMC4-mCherry (I) or TMC5-mCherry (L), resulting in a red signal (C). (M and N)
Specificity of anti-TMC1 antibodies was further validated by lack of label in stereocilia
from Tmc1Δ/ Δ
mice. (O) Specificity of anti-TMC2 antibody was further validated by lack
of label in stereocilia from Tmc2Δ/ Δ
mice. [Scale bars = 5µm]. Related to Figures 2 and 4.
Figure S4. Expression of TMC1-mCherry and TMC2-AcGFP in mature hair
bundles. (A) Confocal image of IHC and OHC stereocilia counterstained with phalloidin
conjugated with AlexaFluor405 (blue), from P10 transgenic mice expressing TMC1mCherry
(red) and TMC2-AcGFP (green). Only TMC1-mCherry fluorescence puncta
(red) are detectable at this age. (Tmc1-mCherry Line 1; Tmc2-AcGFP Line 1). (B)
Confocal image of OHC stereocilia counterstained with phalloidin conjugated with
AlexaFluor488 (green) from transgenic mice expressing TMC1-mCherry (red). TMC1mCherry
fluorescence puncta localized predominantly at stereocilia tips at P15
(Line2). [Scale bar = 5µm]. Related to Figure 5.
Figure S5. Frequency distribution and fluorescence intensity of TMC2 in relation to
their distance to the closest TMC1 puncta in P7 inner hair cells. (A) Histogram
showing the frequency distribution of TMC2-AcGFP puncta centroids according to their
distance to the centroid of the corresponding closest TMC1-mCherry neighboring
punctum (histogram bin size = 30 nm; number of puncta = 650; number of hair cells =12).
(B) Scatter plot of fluorescence intensity (in arbitrary units, AU) of each TMC2-AcGFP
punctum as a function of the distance to their corresponding closest TMC1-mCherry
neighboring punctum. Related to Figure 5.
Figure S6. Analysis of TMC1-mCherry and TMC2-AcGFP localization in cell
bodies is hindered by autofluorescence. (A and B) Two examples of confocal
fluorescence images showing longitudinal cross-sections through IHCs from transgenic
mice expressing TMC1-mCherry and TMC2-AcGFP on Tmc1Δ/Δ
background (Line 4 and
Line2, respectively, P10). The level of fluorescence intensity from hair cell bodies
(asterisk) is comparable to autofluorescence from neighboring support cells (double
asterisks). Granules with strong red and green autofluorescence are observed in
supporting cells (arrowheads). (C-F) Comparison of fluorescence signal from hair cell
body of mice expressing TMC1-mCherry (Line 2, P9) and wild type mice (P9).
Fluorescence signal from TMC1-mCherry expressing hair cells in the cell body (C and D,
asterisk) was observed to be subtly stronger than autofluorescence in wild type hair cells
(E and F, asterisk). Autofluorescent granules were present in transgenic and wild type
cells. [Scale bar = 10µm]. Related to Figure 5.
Supplemental Experimental Procedures
Generation of Transgenic Mice.
The bacterial artificial chromosomes (BACs) MSMG01-526L10 and RP24-293G8,
encoding the mouse Tmc1 and Tmc2 genes, respectively, were purchased from Riken
Bioresource Center DNA Bank and Children’s Hospital Oakland Research Institute,
respectively. MSMG01-526L10 contains the Tmc1 gene, including 6.5-kb upstream and
4.4-kb downstream regions of the gene. In order to confirm that MSMG01-526L10
produced functional TMC1 protein, the unmodified BAC was used to generate a
transgenic mouse line that was backcrossed onto mice with a targeted deletion of Tmc1
(Tmc1Δ
, Kawashima et al. 2011). The progeny had normal hearing (data not shown),
confirming that MSMG01-526L10 contains the regulatory elements required for correct
tissue-specific and temporal expression to rescue the hearing phenotype of Tmc1Δ
mice.
RP24-293G8 contains the Tmc2 gene, in addition to four neighboring genes: Tgm6, Snrpb,
Nop56 and idh3b. In order to fuse fluorophore tags to the C-terminus of TMC1
(Accession # AF417579) and TMC2 (Accession#AF417581), the stop codon of each
gene (c. 2272-2274 for Tmc1 and c. 2665-2667 for Tmc2) was replaced with cDNA
encoding mCherry or AcGFP (Clontech) by recombineering (Gene Bridge). Modified
BACs were purified and injected into (C57BL/6 x SJL) F2 mouse eggs at the University
of Michigan Transgenic Animal Model Core Facility. Founder mice from each BAC,
(Tg(Tmc1/mCherry)1, 2, 3, 4Ajg or Tg(Tmc2/AcGFP)1, 2, 3, 4Ajg) were backcrossed
onto mice segregating targeted deletion alleles of Tmc1 and Tmc2 (Tmc1Δ
and Tmc2Δ
,
(Kawashima et al., 2011). All mice were maintained per the National Institutes of Health
Animal Care and Use Committee.
Genotype Analyses and ABRs
Experiments were conducted in accordance with animal protocols approved by the NIH
Animal Care and Use Committee (Protocol #1264). Founder mice and offspring carrying
the BAC transgenes were identified by PCR analysis of genomic DNA isolated from tail
biopsies. Presence of the full-length BAC was verified using primers recognizing unique
sequences at the ends of each BAC, as well as the mCherry or AcGFP cDNA inserts (See
Table S1). For both Tmc1-mCherry and Tmc2-AcGFP transgenic mice, four founder
lines with the full-length BAC constructs were selected. To genotype the Tmc1Δ
and
Table S1. (Refers to Figure 1-5) Bacterial artificial chromosome PCR primers
# Name Nucleotide Sequence Target Region
1 377
378
5’-GGCCGCTAATACGACTCACTAT
5’-AGTAGAGTCCAGGCATGCTAAAAT
Tmc1-mCherry BAC SP6
2 389
390
5’-GCACAACACATGTTTATTCACTCA
5’-CCGTCGACATTTAGGTGACACTAT
Tmc1-mCherry BAC T7
3 Tmc1mCherryL
Tmc1mCherryR
5’-TTCACTTGCCCTTCTTCATCTC
5’-CGCCCTCGATCTCGAACT
mCherry
4 293G8T7F
293G8T7R
5’-GGTCGAGCTTGACATTGTAGGACT
5’-TGACATGGATAAAGAGTGTGGATCA
Tmc2-AcGFP BAC T7
5 293G8SP6F
293G8SP6R
5’-TGCAGTCGTAAAAGTCAGAACTGTG
5’-AAAGTCTTGTTTAGCTTTGGGCATC
Tmc2-AcGFP BAC SP6
6 Tmc2AcGFPL
Tmc2AcGFPR
5’-ACCATGTTGGGTCTCAACCAC
5’-TGAACTTGTGGCCATTCACAT
AcGFP
Tmc2Δ
alleles, which are originally derived from the strain 129, informative simple
sequence repeats flanking both sides of Tmc1 and Tmc2 were selected (See Table S2).
To estimate BAC transgene copy numbers in each line, quantitative PCR (q-PCR) was
performed with primers corresponding to the deleted exons in Tmc1Δ
and Tmc2Δ
(See
Table S3). Auditory Brainstem Response (ABR) thresholds were measured as described
previously (Noguchi et al., 2006).
Table S2 (Refers to Figure 1-5) Tmc1- and Tmc2-Linked Short Tandem Repeat PCR
Primers
# Name Nucleotide Sequence Strain/Product Size
(bp)
1 Tmc1R5L
Tmc1R5R
5’-CAACACAAGAAGGAGAGGAAA
5’-CTTCCCCTGAAAAACTGTGTC
B6/117
SJL/127
129/123
2 Tmc1L4L
Tmc1L4R
5’-TCCCTCAGTTGGTTGTTTAT
5’-CCATGTGTGCATGTATGTGA
B6/109
SJL/97
129/113
3 Tmc2L12L
Tmc2L12R
5’-CCCCATATGCACACACACAT
5’-TCAAAAGTTGTCCCCTGACC
B6/229
SJL/238
129/233
4 Tmc2RR2L
Tmc2RR2R
5’-TCCCTGTAGTGGGTATCCTCTTT
5’-TCCAGGGACTCTGATCTCTGTAG
B6/188
SJL/191
129/191
Table S3 (Refers to Figures 1-5) qPCR Primers and Probes
Gene Name Nucleotide Sequence
Jun
(Control)
Forward
primer
Reverse
primer
Probe
5’-TCACTATTGGCAACGAGCG
5’-TCAGTTTGGTTCCTCAGTGG
5’-TGTTCCTCA/ZEN/GGTGGATGTACGGC
Tmc1Ex8 Forward
primer
Reverse
primer
Probe
5’-ATGAGGCTGAAGGTCAACAC
5’-TGGATGCAGGGATGATGTTC
5’-TCTGGAAAG/ZEN/CTGTGGCGTGATGG
Tmc2Ex7 Forward
primer
Reverse
primer
Probe
5’-AAGCCCCATCAACACTATCTG
5’-GCATCTTACTTCATCTTTCTCCG
5’-
ACAGGGAAT/ZEN/TCTGTTCTTACCTCTGGGA
	  
Hair Cell Electrophysiology
Animals were decapitated using methods approved by Animal Care and Use Committee
of Boston Children’s Hospital (Protocol #1959 and #2146). Organs of Corti were
dissected from postnatal day P3 and P7 mice, mounted in recording chambers and viewed
on an Axio Examiner.A1 upright microscope (Carl Zeiss) equipped with a 63x waterimmersion
objective and differential interference contrast (DIC) optics. In all
preparations, the tectorial membrane was peeled off the tissue. Electrophysiological
recordings were performed at room temperature (22 °C-24 °C) in standard solutions
containing (in mM): 140 NaCl, 5.8 KCl, 10 HEPES, 0.7 NaH2PO4, 1.3 CaCl2, 0.9 MgCl2,
and 5.6 D-glucose, vitamins (Invitrogen, 1:100) and amino acids (Invitrogen, 1:50) as in
in MEM (Invitrogen, pH 7.4; 310 mOsm/kg). Recording electrodes (3-4 MΩ) were
pulled from R-6 glass (King Precision Glass) and filled with (in mM): 140 CsCl, 5
EGTA-KOH, 5 HEPES, 2.5 Na2ATP, 3.5 MgCl2, and 0.1 CaCl2 (pH 7.4; 284 mOsm/kg).
The whole-cell, tight-seal technique was used to record mechanotransduction currents
with an Axopatch 200B (Molecular Devices). Cells were held at –84 mV. Currents were
filtered at 5 kHz with a low-pass Bessel filter, digitized at >20 kHz with a 12-bit
acquisition board (Digidata 1440A), and recorded using pClamp 10 software (Molecular
Devices). Hair bundles of IHCs and OHCs were deflected using stiff glass probes
mounted on a PICMA chip piezo actuator (Physik Instruments) driven by an LPZT
amplifier (Physik Instruments) and filtered with an 8-pole Bessel filter at 40 kHz to
eliminate residual pipette resonance as previously described (Stauffer and Holt, 2007).
Pipettes were designed to fit into the concave aspect of the array of hair cell stereocilia
for whole-bundle recordings. An apical extracellular perfusion protected the hair bundles
from intracellular solution with rate of 0.08 ml/min using pipettes with tip sizes 50-100
µm. Current traces were imported and analyzed using Origin 8.0 (OriginLab).
Hair Cell Electrophysiology for reverse-polarity currents
Experiments were conducted in accordance with animal protocols approved by the NIH
Animal Care and Use Committee (Protocol #1215). Organs of Corti were dissected from
postnatal day P4 to P8 mice, placed in a recording chamber mounted on an upright
microscope (Olympus), and visualized using a 63x 0.9 NA water-immersion objective
and infrared differential interference contrast. To determine the distance along the
cochlea, the total length of the organ of Corti (equal to 1) was virtually segmented and
longitudinal distances from the apex were divided by the total length of the cochlea,
similar to an approach reported previously (Kim and Fettiplace, 2012). Recordings were
made from hair cells located 10 – 50 % from the apex of the cochlea. Regions 10 – 30%
from the apex of the cochlea were considered to be apical, regions 30 – 50 % from the
apex of the cochlea are referred to as mid-apical. Extracellular solution used to dissect
and maintain the tissue during recordings contained (in mM): 140 NaCl, 5.6 KCl, 10
HEPES, 0.7 NaH2PO4, 1.3 CaCl2, 0.9 MgCl2, 5.6 D-glucose, 0.4 Na-ascorbate, 2 Napyruvate
(pH 7.4; 305-310 mOsm/kg). Recording electrodes (2.5-4 MΩ) pulled from
thick-walled borosilicate glass (Sutter Instruments) were filled with intracellular solution
containing (in mM): 125 CsCl, 1 EGTA, 10 HEPES, 3.5 MgCl2, 5 Mg-ATP, 3 Naascorbate,
5 Tris-phosphocreatine (pH 7.3; 290-295 mOsm/kg). Data were filtered at 10
kHz using a Multiclamp 700B amplifier (Molecular Devices) and sampled at 50 kHz.
Cells were held at -80 mV, no correction for the liquid junction potential was applied.
Series resistance (4-10 MΩ) was compensated by 80-90 % and experiments in which the
series resistance increased by >20% were excluded from further analysis. Hair bundles
were mechanically stimulated with a fluid jet system (HSPC-1, ALA Scientific)
connected to a glass pipette (tip diameter 8-15 µm). The position of the stimulating
pipette was adjusted to elicit a maximal MET current. The stimulus used was a 40-Hz
sinusoidal wave generated in IgorPro 6 (WaveMetrics), driver voltage inducing maximal
control MET currents were ±1.5-2.5 V. In experiments with BAPTA treatment, a 100
mM stock solution was made by adding adequate amount of BAPTA into 0.3 N sodium
bicarbonate dissolved in extracellular solution. The stock was then added to our bath
solution to obtain final concentration of 5 mM BAPTA. This solution was either bath
applied using gravity fed application system at a perfusion rate of 2-3 ml/min and reached
the recording chamber with the tissue in approximately 2 minutes after switching from
regular, BAPTA-free solution, or locally applied to the hair bundle of the recorded hair
cell using a patch-clamp pipette connected to a Picospritzer (Parker Hannifin, Pine Brook,
NJ). All data were acquired and analyzed using custom routines written in IgorPro 6
(WaveMetrics). All experiments were performed at room temperature.
Detection of TMC1-mCherry and TMC2-AcGFP
Temporal bones from mice were rapidly (1-2 min) isolated and placed in Leibovitz’s L-
15 (L-15) medium supplemented with 2 mM CaCl2 and 10 mM HEPES, and immediately
fixed in 4 % PFA in L-15 medium for 20 min. The samples were washed with PBS, and
the organ of Corti, saccule and utricle were excised. Samples were then permeabilized
with 0.5 % Triton X-100 in PBS containing Alexa Fluor 405, 488 or 568 phalloidin for
30 min before washing and mounting on a microscope slide with water. For FM1-43
uptake, the cochleae from P10 mice were excised and 1 ml of N-(3triethylammoniumpropyl)-4-(4-(dibutylamino)styryl)pyridinium
dibromide (FM1-43FX,
Invitrogen) at 5 μM in L-15 medium at room temperature was applied to the cochlear
duct from oval and round windows within 1 min, then, washed three times with 1 ml of
L-15 medium, and fixed in 4 % PFA for 20 min at room temperature. The organ of Corti
were excised and mounted for imaging. Unless specified we focused our imaging and
analyses to the mid-apical region of the organ of Corti.
Immunohistochemistry
Polyclonal antibodies against mouse TMC1 (PB277 and PB612) and mouse TMC2
(PB361) were developed (Princeton Biomolecules and Covance) in rabbits immunized
with synthetic peptides (PB277: CDEETRKAREKERRRRLRRGA; PB612:
CAVKRSQEFAQQDPDTLG; PB361, CGSQPPRGRRDSGQPQSQTY; see also Table
S4). After affinity purification of the antibodies against their corresponding immunizing
peptides, their specificity was verified in COS7 cells transfected with TMC1- or TMC2mCherry
(Figure S3). Selectivity of the antibodies was also tested against mCherrytagged
TMC4 (TMC4-mCherry), recently shown to be expressed in hair cells (Liu et al.,
2014), as well as TMC5-mCherry (Figure S3). Rats and mice were euthanized in
accordance to National Institutes of Health (NIH) guidelines, and their temporal bones
fixed by immersion in 4 % paraformaldehyde (PFA) in phosphate buffered saline (PBS;
pH 7.4) for 20 min at room temperature. Inner ear (cochlear and vestibular) sensory tissue
was dissected in PBS, permeabilized with 0.5 % Triton X-100 for 30 min and blocked
overnight at 4 °C with 4% bovine serum albumin (BSA) in PBS. Tissue was then
incubated with primary antibody for 2 h, rinsed with PBS, stained with Alexa Fluor 488or
568- conjugated secondary antibody (Life Technologies) for 1 h, counterstained with
0.001 U/µl Alexa Fluor 488- or 568-phalloidin (Molecular Probes), and mounted using
Prolong Gold Antifade (Molecular Probes). Specificity of anti-TMC1 antibodies was
further validated by lack of label in stereocilia from Tmc1Δ/ Δ
mice (Figure S3 M and N).
Specificity of anti-TMC2 antibody was further validated by lack of label in stereocilia
from Tmc2Δ/ Δ
mice (Figure S3 O).
Table S4 (Refers to Figures 2 and 4) Peptide sequences used for generation of
antibodies.
Name Antigen Peptide Sequence Amino acid positions
PB277 DEETRKAREKERRRRLRRGAE mouse TMC1 aa52-73
PB612 AVKRSQEFEFAQQDPTLG mouse TMC1 aa380-396
PB361 GDQPPRGRRDSGQPQSPTY mouse TMC2 aa855-873
Microscopy
Microscopy was performed using a Nikon TiE inverted fluorescence microscope,
outfitted with a spinning disk confocal scan head (Yokogawa), 100x Apo TIRF 1.49 N.A.
objective, and Andor DU-888 EM-CCD. The effective pixel size at the camera sensor
was 30 nm, enabled by a 1.5x tube lens and a total of 3x secondary magnification.
Although such pixel dimensions exceeded the resolving ability of the objective as defined
by its NA, it allowed for the high dynamic range inherent to datasets acquired with long
exposure times to be effectively spread, as shown previously (Burnette et al., 2011). NISElements
imaging software was utilized for image acquisition and analysis.
	  
Supplemental	  References
	  
Burnette,	  D.T.,	  Sengupta,	  P.,	  Dai,	  Y.,	  Lippincott-­‐Schwartz,	  J.,	  and	  Kachar,	  B.	  (2011).	  
Bleaching/blinking	  assisted	  localization	  microscopy	  for	  superresolution	  imaging	  
using	  standard	  fluorescent	  molecules.	  Proceedings	  of	  the	  National	  Academy	  of	  
Sciences	  of	  the	  United	  States	  of	  America	  108,	  21081-­‐21086.	  
Kawashima,	  Y.,	  Geleoc,	  G.S.,	  Kurima,	  K.,	  Labay,	  V.,	  Lelli,	  A.,	  Asai,	  Y.,	  Makishima,	  T.,	  Wu,	  
D.K.,	  Della	  Santina,	  C.C.,	  Holt,	  J.R.,	  et	  al.	  (2011).	  Mechanotransduction	  in	  mouse	  inner	  
ear	  hair	  cells	  requires	  transmembrane	  channel-­‐like	  genes.	  J	  Clin	  Invest	  121,	  4796-­‐
4809.	  
Liu,	  H.,	  Pecka,	  J.L.,	  Zhang,	  Q.,	  Soukup,	  G.A.,	  Beisel,	  K.W.,	  and	  He,	  D.Z.	  (2014).	  
Characterization	  of	  transcriptomes	  of	  cochlear	  inner	  and	  outer	  hair	  cells.	  The	  
Journal	  of	  neuroscience	  :	  the	  official	  journal	  of	  the	  Society	  for	  Neuroscience	  34,	  
11085-­‐11095.