The ?10 nicotinic acetylcholine receptor subunit
is required for normal synaptic function and
integrity of the olivocochlear system
Douglas E. Vetter*†, Eleonora Katz‡§, Ste ´phane F. Maison¶?, Julia ´n Taranda*‡, Sevin Turcan**, Jimena Ballestero‡,
M. Charles Liberman¶§, A. Bele ´n Elgoyhen‡††, and Jim Boulter‡‡
*Department of Neuroscience, Tufts University School of Medicine, Boston, MA 02111; **Department of Biomedical Engineering, Tufts University, Boston,
MA 02111;‡Instituto de Investigaciones en Ingenierı ´a Gene ´tica y Biologı ´a Molecular (INGEBI), Consejo Nacional de Investigaciones Cientı ´ficas y Te ´cnicas
(CONICET), Buenos Aires 1428, Argentina;§Departamento de Fisiologı ´a, Biologı ´a Molecular y Celular, Facultad de Ciencias Exactas y Naturales, Universidad
de Buenos Aires, Buenos Aires 1428, Argentina;¶Eaton-Peabody Laboratory, Massachusetts Eye and Ear Infirmary, Boston, MA 02114;?Department of
Otology and Laryngology, Harvard Medical School, Boston, MA 02115;††Departamento de Farmacologı ´a, Facultad de Medicina, Universidad de
Buenos Aires, Buenos Aires 1121, Argentina; and‡‡Department of Psychiatry and Biobehavioral Sciences, Hatos Research Center for
Neuropharmacology, Brain Research and Molecular Biology Institutes, University of California, Los Angeles, CA 90024
Edited by David Julius, University of California, San Francisco, CA, and approved November 7, 2007 (received for review September 10, 2007)
Although homomeric channels assembled from the ?9 nicotinic
acetylcholine receptor (nAChR) subunit are functional in vitro,
electrophysiological, anatomical, and molecular data suggest that
native cholinergic olivocochlear function is mediated via hetero-
meric nAChRs composed of both ?9 and ?10 subunits. To gain
insight into ?10 subunit function in vivo, we examined olivo-
cochlear innervation and function in ?10 null-mutant mice. Elec-
trophysiological recordings from postnatal (P) days P8–9 inner hair
cells revealed ACh-gated currents in ?10?/?and ?10?/?mice, with
no detectable responses to ACh in ?10?/?mice. In contrast, a
proportion of ?10?/?outer hair cells showed small ACh-evoked
currents. In ?10?/?mutant mice, olivocochlear fiber stimulation
failed to suppress distortion products, suggesting that the residual
?9 homomeric nAChRs expressed by outer hair cells are unable to
an abnormal olivocochlear morphology and innervation to outer
hair cells and a highly disorganized efferent innervation to the
inner hair cell region. Our results demonstrate that ?9?/?and
?10?/?mice have overlapping but nonidentical phenotypes. More-
over, ?10 nAChR subunits are required for normal olivocochlear
activity because ?9 homomeric nAChRs do not support mainte-
nance of normal olivocochlear innervation or function in ?10?/?
cochlea ? electrophysiology ? inner hair cells ? outer hair cells
unique subset of cells that respond to mechanical cues. These
hair cells possess apical mechanoreceptors and specialized ba-
solateral membranes that act in concert to transduce mechanical
stimuli into electrical signals (1). In mammals, cochlear hair cells
are anatomically and functionally divided into inner and outer
hair cells (IHCs and OHCs, respectively). IHCs are responsible
for transducing acoustic stimuli and exciting the fibers of the
cochlear nerve, whereas OHC are involved in the mechanical
amplification and fine tuning of cochlear vibrations via their
electromotile response (2, 3).
Both OHCs and type-I spiral ganglion cell processes receive
descending cholinergic innervation, which originates in the
superior olivary complex (4). Although the precise role of
the olivocochlear (OC) system in hearing remains uncertain, the
effects of activating efferent terminals forming synapses with
OHCs have been well described (5–7). Acetylcholine (ACh), the
principal neurotransmitter released by OC terminals (8), binds
ing to calcium influx, activation of small-conductance calcium-
activated potassium channels, and subsequent hair cell hyper-
polarization (9–17). As with electrical stimulation of the
he sensory epithelia responsible for hearing (cochlea) and
balance (saccule, utricle, and cristae ampullaris) share a
olivocochlear bundle (10), the result of OHC hyperpolarization
is to reduce auditory afferent output via suppression of basilar
membrane motion (18).
Combined immunohistochemical (19, 20), electrophysiologi-
cal (9, 21–23), molecular biological (7, 24), and in situ hybrid-
ization studies (21, 25–27) suggest that the nAChR subtype
present at efferent hair cell synapses is assembled from both ?9
ion channels, ?9 subunits form functional homomeric nAChRs
in Xenopus oocytes (25). Importantly, coinjection of cRNAs
encoding both the ?9 and ?10 subunits results in an ?100-fold
increase in the amplitude of ACh-gated currents, and the
resulting heteromeric ?9?10 nAChRs possess the distinctive
pharmacological and biophysical properties of native hair cell
cholinergic receptors (21, 28).
Composed exclusively of ? subunits, the hair cell ?9?10
nAChR subtype is unusual. Moreover, the subunit composition
and heterologous expression data raise some interesting ques-
tions. For example, do hair cell ?9 homomeric nAChRs function
in vivo (as they do in vitro)? If so, could they support normal hair
cell cholinergic biology? If not, what added functionality is
contributed by the ?10 subunit? Is expression of the ?10 subunit
essential for normal hearing or proper efferent innervation of
hair cell synapses? Is the ?10 protein required to obtain a full
complement of OC efferent effects in vivo? To gain insight into
these questions and examine the role of ?10 subunits in mam-
malian hair cells, we engineered a strain of mice that harbors a
null mutation in the Chrna10 gene. Here, we report that even
though a proportion of OHCs remain minimally responsive to
ACh (because of the presence of residual ?9 homomeric recep-
nAChRs are required for both normal efferent activation of
these hair cells and for development (or maintenance) of normal
OC synapse structure and function.
Hair Cell Electrophysiology. Given that expression levels of genes
important for efferent function were generally unaltered in
Author contributions: D.E.V. and E.K. contributed equally to this work; D.E.V., E.K., S.F.M.,
A.B.E., and J. Boulter designed research; D.E.V., E.K., S.F.M., J.T., S.T., J. Ballestero, and J.
Boulter performed research; D.E.V., E.K., S.F.M., J.T., S.T., J. Ballestero, M.C.L., A.B.E., and
J. Boulter analyzed data; and D.E.V., E.K., and A.B.E. wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
†To whom correspondence should be addressed at: Tufts University School of Medicine,
Department of Neuroscience, 136 Harrison Avenue, Boston, MA 02111. E-mail:
This article contains supporting information online at www.pnas.org/cgi/content/full/
© 2007 by The National Academy of Sciences of the USA
December 18, 2007 ?
vol. 104 ?
?10?/?mice save for the ?10 gene itself [supporting information
(SI) Table 1 and SI Methods], and that nAChR ?9 subunits form
functional homomeric channels in Xenopus oocytes (21, 25), we
sought to further investigate the functional state of the hair cells
in the ?10?/?mice, with special emphasis on determining whether
currents attributable to ?9 homomers could be identified.
We compared cholinergic responses of IHCs isolated from
?10?/?, ?10?/?, and ?10?/?mice at postnatal (P) days P8–9, a
terminals (29, 30) and robustly respond to ACh. Inward currents
were elicited at ?90 mV by the application of 100 ?M ACh to
IHCs from both ?10?/?(?200 ? 48 pA; n ? 3 of 3 cells tested)
and ?10?/?(?517 ? 50 pA, n ? 3 of 3 cells tested) mice (Fig.
1A). In contrast, no response to either 100 ?M or 1 mM ACh
(n ? 0 of 7 cells tested) was detected in IHCs from ?10?/?mice.
IHCs from both ?10?/?and ?10?/?mice also exhibited outward
currents at ?40mV (Fig. 1B; wild type: 236 ? 44 pA; n ? 3 of
3 cells tested; heterozygous: 278 ? 80 pA; n ? 3 of 3 cells tested),
indicating functional coupling to SK channels (9, 28), whereas no
response was found in IHCs from ?10?/?mice (n ? 0 of 8 cells
To determine whether IHCs respond to synaptic release of
mM KCl to depolarize the efferent terminals, thus increasing the
frequency of ACh release (9, 28). KCl depolarization generated
ACh-inducible synaptic currents in IHCs from both ?10?/?(n ?
3 of 3 cells tested) and ?10?/?(n ? 3 of 3 cells tested) mice (Fig.
1C Left and Center) but not in IHCs from ?10?/?mice (n ? 0 of
8 cells tested; Fig. 1C Right). Moreover, even when adding either
100 ?M or 1 mM ACh in the presence of 40 mM K?(Fig. 1C),
a procedure that uncovers hair cells with small responses to ACh
due to the change in the K?equilibrium potential and the
concomitant increase in the driving force for K?ions at the hold-
ing voltage (?90 mV) (9), no responsive IHCs were observed.
both ?10?/?(?103 ? 24 pA; 6 of 7 cells tested) and ?10?/?
(?76 ? 40 pA; 3 of 4 cells tested) mice (Fig. 2A). In ?10?/?, only
1 of 11 cells tested responded to 1 mM ACh (inward current of
?13 pA). To determine whether ACh-evoked responses were
coupled to activation of SK channels, OHCs were voltage-
clamped at ?40 mV and perfused with 1 mM ACh (Fig. 2B).
OHCs from ?10?/?(12 of 13 cells tested) and ?10?/?(16 of 17
cells tested) exhibited robust outward currents (104 ? 46 and
171 ? 37 pA, respectively). In ?10?/?, consistent with the lack
of ACh responses in most of the cells tested at ?90 mV, only 2
of 24 OHCs responded to ACh with an outward current (am-
plitudes of 30 and 58 pA). In 40 mM KCl (Fig. 2C), OHCs from
both ?10?/?and ?10?/?mice exhibited synaptic currents (8 of
13 and 15 of 16 cells, respectively). Synaptic currents were not
2C). After application of 1 mM ACh plus 40 mM KCl (Fig. 2D),
the number of responsive OHCs increased: 10 of 11 in ?10?/?
and 15 of 15 in ?10?/?(range from ?160 to ?1,300 pA). In
contrast to what was observed in IHCs, boosting the system by
elevating external K?revealed small inward currents (?40 to
?151 pA) in 11 of 24 OHCs from ?10?/?mice, highly suggestive
of ?9 subunits assembling into functional homomeric receptors.
Both ?9 homomeric and ?9?10 heteromeric nAChRs can be
distinguished from other nAChR subtypes by their unusual
pharmacological profiles (21, 25, 31). In particular, activation by
ACh but lack of activation by nicotine is a hallmark of receptors
assembled from ?9 and ?10 subunits (21, 25, 31, 32). As shown
in Fig. 3A, 300 ?M nicotine did not evoke currents in OHCs that
were responsive to ACh from ?10?/?mice (n ? 0 of 3 cells
tested). Considered together, our quantitative RT-PCR (SI
Table 1) and electrophysiological data suggest that ACh-evoked
currents observed in Chrna10?/?OHCs are mediated by homo-
meric ?9 nAChRs. The lack of comparable activity in Chrna9?/?
mice (Fig. 3C; n ? 0 of 14 cells tested) further supports this
Cochlear Function. Because OC feedback can alter cochlear
thresholds and is required for development of normal cochlear
function and morphology (6, 7), we investigated baseline co-
chlear response sensitivity in ?10?/?mice. No differences in
threshold or suprathreshold responses (SI Fig. 6) were detected
in either auditory brainstem responses (ABRs), the summed
activity of auditory neurons evoked by short tone pips or
distortion product otoacoustic emissions (DPOAEs), which are
distortions of the sound input created and amplified by normally
functioning OHCs and actively transduced out of the cochlea
and measured in the ear canal.
To test the effects of ?10 gene deletion on OC function in vivo,
we measured the effects of OC electrical activation on DPOAE
amplitudes. DPOAE amplitudes in wild-type mice always show
a fast suppression after OC fiber stimulation that is visible in the
first DPOAE measurement after shock train onset (Fig. 4A).
Suppression is greatest for mid-frequency tones (Fig. 4B), mir-
roring the peak of OHC efferent terminal density in the middle
of the cochlear spiral (33).
?10?/?mice. (A) Effects of ACh at Vholdof ?90 mV in the three genotypes. (B)
Same as A at a Vholdof ?40 mV. Note that in A and B, no responses could be
cells with high potassium (40 mM at Vholdof ?90 mV) causes a change in the
there are synaptic currents appearing on top of the holding current because
upon challenging the cells with either 100 or 1,000 ?M ACh. Results are
representative of those obtained in three IHCs from ?10?/?mice, three IHCs
from ?10?/?mice, and eight IHCs from three ?10?/?mice. Holding currents at
?90 mV ranged from ?100 to ?200 pA and from ?200 to ?400 pA in 5.8 and
40 mM K?, respectively. At ?40 mV in 5.8 mM K?, holding currents ranged
from 0 to 200 pA.
IHC whole-cell recordings from P8–9 N5 B6.Cast ?10?/?, ?10?/?, and
Vetter et al.
December 18, 2007 ?
vol. 104 ?
no. 51 ?
In an earlier study of ?9?/?mice, we used a paradigm in which
DPOAEs were alternately measured with and without shocks in
repeating 6-second trial intervals. In this paradigm, only sup-
pressive effects of olivocochlear stimulation are seen in ?9?/?
mice, and all these suppressive effects are eliminated by ?9 gene
deletion (7). Since that time, we modified our paradigm to one
in which the shock train is maintained for 70 seconds and
DPOAEs measured before, during, and after this shock epoch.
This paradigm reveals a robust postshocks enhancement of
DPOAE amplitudes in both ?9?/?and ?9?/?mice (Fig. 4A)
(also see ref. 34). In contrast to the fast-onset suppression, slow
postshocks enhancement tends to increase monotonically in
amplitude with increasing stimulus frequency (Fig. 4B).
In ?10?/?mice, efferent-evoked suppression of DPOAEs was
never observed (n ? 12 mice tested, each ear tested at six
different DPOAE-evoking frequencies). Interestingly, a during-
(7 of 10 ears tested that met DPOAE threshold criteria), which,
after shock-train offset, behaved similarly to the normal post-
shocks enhancement observed in wild types both with respect to
its amplitude and offset time constant (Fig. 4A). This result is
similar to that observed in ?9?/?(34) (reassessed and included
in Fig. 4 for comparison).
Cochlear Morphology. In the OHC region, efferent innervation in
wild-type mice consists of clusters of synaptic terminals under
the three rows of OHCs, except at the apical extreme of the
cochlea. Similar to observations in ?9?/?mice (7), efferent
terminals contacting OHCs of ?10?/?mice were larger in size
but fewer in number (SI Fig. 7). The terminals measured an
average 2.90 ?m in diameter (?0.09 ?m), ?20% larger than
those from the same cochlear region of wild-type mice (2.38 ?
0.07 ?m; t test, P ? 0.00001). Although OHCs of ?10?/?mice
were contacted by abnormally large efferent terminals, and were
fewer in number than wild types, the number of those terminals
under each OHC was greater than that in ?9?/?mice (7). A
count of 100 random OHCs from contiguous 250-?m regions of
the middle turns of three ?10?/?mice revealed an equal
probability of the OHCs possessing either one or multiple
efferent terminals (50% occurrence for each case) whereas 67%
of synaptic contacts with OHCs in ?9?/?mice consisted of single
boutons. Only row three of ?10?/?mice possessed a statistically
significant number of single terminals (P ? 0.01) compared with
multiply innervated hair cells within the same row of wild-type
mice. However, unlike ?10?/?mice, OHCs of the null-mutant
?10?/?, and ?10?/?mice. (A) Representative records of the effects of 1 mM
ACh at a Vholdof ?90 mV in the three genotypes (positive/studied cells ? 6/7
?10?/?; 3/4 ?10?/?; 1/11 ?10?/?). (B) Same as A at a Vholdof ?40 mV (positive/
studied cells ? 12/13 ?10?/?; 16/17 ?10?/?; 2/24 ?10?/?). (C) Representative
at Vholdof ?90 mV). A change in holding current due to the change in the K?
?10?/?mice (Left and Middle, respectively), there are synaptic currents ap-
pearing on top of the holding current because of the release of ACh from
depolarized efferent terminals (8 of 13 OHCs and 15 of 16 OHCs from ?10?/?
and ?10?/?mice, respectively). No synaptic activity can be observed in OHCs
from ?10?/?mice (Right; n ? 24 cells). (D) After boosting the system by
increasing the K?driving force, almost all OHCs studied from both the ?10?/?
and ?10?/?mice (Left and Middle, respectively) showed ACh-evoked re-
sponses (Vholdof ?90 mV; positive/studied cells ? 10/11 ?10?/?, 15/15 ?10?/?).
Interestingly, OHCs from ?10?/?mice also exhibited small but consistent
inward currents in response to ACh (11 of 24 cells studied, Right and Inset).
Holding currents at ?90 mV ranged from ?100 to ?300 pA and from -?250
to ?600 pA in 5.8 and 40 mM K?, respectively. At ?40 mV in 5.8 mM K?,
holding currents ranged from 0 to 200 pA.
OHC whole-cell recordings from P10–13 N5-N8 B6.Cast ?10?/?,
apical turns) excised from P10–13 ?10?/?and ?9?/?mice. (A) In OHCs from
?10?/?mice, no responses to 300 ?M nicotine (Left and Inset) could be
obtained in the same OHCs in which 1 mM ACh elicited an inward current
(Middle and Inset; n ? 3). (B) Representative record of the lack of effect of 1
mM ACh applied in the presence of 40 mM KCl in OHCs from ?9?/?mice (n ?
14 cells). Vholdof ?90 mV in both A and B.
Whole-cell recordings in OHCs from cochlear preparations (medio-
www.pnas.org?cgi?doi?10.1073?pnas.0708545105 Vetter et al.
mice contacted by three or more efferent terminals were ex-
In wild-type mice, the inner spiral bundle contains efferent
fibers synapsing with cochlear nerve dendrites in the region
below and surrounding the IHCs. Most immunostained termi-
nals are found on the modiolar side of the IHC (closer to the
nerve trunk), but a lesser number of smaller and less brightly
stained efferent boutons are found on the pillar side of the IHCs
as well (Fig. 5A, arrowheads). The synaptophysin-stained inner
spiral bundle of ?10?/?mice appeared disorganized. Modiolar-
side terminals were larger in ?10?/?mice than in wild type, and
there was a paucity of terminals on the pillar side of the hair cell
(Fig. 5B). The degree of disorganization was quantified via a
nearest-neighbor distance analysis of 388 terminals in ?10?/?
mice and 425 terminals in ?10?/?mice (Fig. 5 D and E). The
mean distance between terminals was larger in ?10?/?mice
compared with ?10?/?mice (1.48 ?m ? 0.04 vs. 1.19 ?m ? 0.03
?m, Fig. 5D; unpaired two-tailed t test, P ? 6.7 ? 10?7). The
maximal spread of the nearest-neighbor distance in ?10?/?mice
was increased by ?50% over that of wild type (7 ?m vs. 4.7 ?m),
and the number of terminals with nearest neighbors between 1.7
and 4.7 ?m (the bulk of the terminals above the mean of the
?10?/?mice) was larger in ?10?/?mice (n ? 81) compared with
?10?/?mice (n ? 49). A similar examination of the ?9?/?mice
(656 terminals counted over a similar region as above) revealed
no statistically significant difference in nearest-neighbor dis-
tance with the wild-type mice [1.19 ?m ? 0.03 (WT) vs. 1.13
?m ? 0.02 (KO); Fig. 5 C–E].
Our data illustrate that the ?10?/?phenotype is distinct from
that observed in the ?9?/?mouse line in terms of hair cell
physiological function and synaptic structure. In addition, our
data demonstrate that the residual functional ?9 nAChRs ex-
pressed in ?10?/?mice are insufficient to drive normal OC
efferent function. Thus, our data definitively establish the re-
quirement for ?10 subunits in forming biologically relevant hair
Homomeric ?9 nAChRs reconstituted in Xenopus oocytes
produce small ACh-evoked currents (25). The presence of small
ACh-evoked currents in some ?10?/?OHCs suggests the con-
tinued expression of functional ?9 receptors that likely consist of
homomeric subunits. Lack of nicotine-induced activation in
OHCs that are otherwise ACh-responsive is consistent with the
presence of ?9 homomeric receptors. Moreover, the fact that
OHCs from ?9?/?mice do not present ACh-evoked currents
rules out the possibility that in the absence of functional ?9?10
nAChRs, small residual muscarinic currents could be disclosed.
Because both ?9 homomeric and ?9?10 heteromeric receptors
have a high Ca2?permeability (35, 36), one might expect
coupling of ?9 nAChRs to an SK2 channel in OHCs of the
?10?/?mice. The fact that outward currents were observed in
OHCs of ?10?/?mice at ?40 mV suggests coupling to a
potassium channel. It is well recognized that nAChRs are
coupled to SK2 channels in wild-type hair cells (9–11, 13–17, 28),
and we show here that ?10?/?mice express normal levels of SK2
transcripts. Thus, it seems reasonable to propose SK2 channels
as the source of the potassium current. The fact that no synaptic
either ?9 homomeric receptors are extrasynaptic, or that cur-
rents are simply too small to detect, perhaps owing to less
efficient insertion into the membrane.
Our inability to detect ACh-evoked responses in IHCs of
?10?/?mice is consistent with the observation that loss of ?10
(but not ?9) transcripts after the onset of hearing is correlated
with the absence of functional ACh receptors in IHCs of
wild-type mice (9, 21). This natural loss of ACh-inducible
response reinforces the interpretation of the experimental data
that ?10 is a key component of functional nAChRs present in
IHCs (9, 28) and suggests either that the number of homomeric
?9 receptors is too small to generate detectable currents, or that
in IHCs, a population of homomeric ?9 receptors is not assem-
bled or inserted into the membrane.
As reported for ?9?/?mice (7), loss of ?10 has no effect on
cochlear baseline sensitivity. This loss is not unexpected given
that there is no sound-evoked activity in the OC fibers at
threshold levels and little spontaneous activity (37). The mam-
malian olivocochlear system constitutes a sound-evoked reflex
pathway that is excited by sound in either ear (38, 39). Electric
activation of the OC system evokes a fast-onset decrease in
cochlear sensitivity, as measured either via afferent responses (5,
7, 40), hair cell receptor potentials (41, 42), DPOAEs (43, 44),
or basilar membrane motion (18). The fact that no fast-onset
DPOAE suppression was observed in the ?10?/?mice indicates
that calcium currents through residual ?9 nAChRs are not
sufficient to drive normal OC activity. Enhancement of
DPOAEs (as seen in both ?9?/?and ?10?/?mice) suggests that
this phenomenon is a normal part of the OC response whose
response enhancement, which slowly decays back to baseline after shock offset. (A) Sample runs of the olivocochlear efferent assay from a wild-type, an ?10?/?,
and an ?9?/?mouse. DPOAE amplitudes are repeatedly measured before, during, and after a 70-second train of shocks to the olivocochlear bundle at the floor
of the IVth ventricle. DPOAE amplitudes are normalized to the mean preshocks value in each case. The dashed ellipses indicate the time windows during which
efferent-evoked effects are sampled to produce the mean data in B. (B) Mean (?SEM) effects of olivocochlear stimulation in groups of ?10?/?vs. ?10?/?ears.
the postshocks value is defined by averaging the 7th to 12th points after shock-train offset.
Olivocochlear efferent function. Deletion of either the ?9 or the ?10 nAChR eliminates suppressive olivocochlear effects, leaving only a during-shocks
Vetter et al.
December 18, 2007 ?
vol. 104 ?
no. 51 ?
initial activity has been unmasked by the loss of the normal,
suppressive cholinergic activity. The molecular mechanism gen-
erating the enhancement remains to be determined but is clearly
not generated by ?9?10-containing nAChRs. It is likely that
additional mechanisms will need to be explored, given that OC
terminals also express GABA and CGRP, and that ACh can also
activate OHCs via muscarinic cholinergic receptors (for a more
complete discussion of these alternatives, see ref. 34).
We previously reported changes in OC innervation to OHCs
in the ?9?/?mouse (7). The dysmorphology of efferent inner-
vation in the ?10?/?mice suggests that the ?10 subunit also plays
a significant role in either the development or homeostatic
their cochlear targets, and, furthermore, that any residual ACh-
evoked activity from activation of ?9 homomeric receptors is not
program. However, the observed synaptic abnormalities ob-
served in ?10?/?mice differ from that of the ?9?/?mice,
possibly because of residual ?9 nAChR activity in ?10?/?mice.
Indeed, whereas OHC terminals of the ?9?/?and ?10?/?mouse
lines are both hypertrophied and abnormal in their number,
?10?/?mice show a greater number of boutons contacting
individual OHCs than do the ?9?/?mice.
Comparing the disorganized efferent innervation to the IHC
region of ?10?/?mice with the more regular pattern of inner-
vation in wild-type and ?9?/?mice suggests that loss of the ?10
gene is more detrimental to synapse formation than elimination
of the ?9 gene, which we have shown completely silences the
cholinergic synapse (ref. 7 and present results). It is unknown
whether homomeric ?9 nAChRs are active in IHCs of ?10?/?
mice (albeit abnormally) during very early postnatal stages or
embryonic development. If this is the case, such activity may
result in a different response to ACh-induced activity compared
indicate either a subtle change in activity that is lost before our
date of examination by electrophysiology (P8), or that the hair
cell nAChRs play a structural or metabolic role in synapse
formation within the IHC region not previously appreciated.
Our data indicate that, although tempting to view ?10 as a
‘‘modulatory’’ subunit given the homomeric ?9 responses elic-
ited in heterologous expression systems (21, 25), ?10 is in fact
absolutely necessary for proper nAChR activity induced by
olivocochlear neurotransmission. It is thus possible to speculate
that ?10 has evolved to serve a special role in mammalian
audition. From the present results, it is clear that although
functional, homomeric ?9 nAChRs are insufficient either in
number or activity to suppress DPOAE amplitudes, and that the
inner ear requires ?9?10 nAChRs to permit CNS modulation of
cochlear mechanics, thereby invoking the physiological roles of
the OC system (e.g., protection from moderate noise-induced
trauma, establishing attention to specific signals, etc.). An evo-
lutionarily diverged ?10 subunit capable of assembling with ?9
and conveying new properties to the nAChR is therefore ap-
parently required to obtain classical OC efferent effects. Our
results complement previous phylogenetic and evolutionary
analysis (45) of the ?10 subunit indicating that ?10 has likely
evolved to give the auditory system a feedback control capability
over the coevolved somatic electromotility (2, 46) that is not
required in nonmammalian species.
Genetic Engineering and Genotyping of ?10 Null-Mutant Mice. Standard pro-
The Chrna10 null-mutant allele has been backcrossed (n ? 5) and maintained
in homozygous congenic B6.CAST-ahl?mice (stock number 002756; Jackson
the National Institutes of Health Guide for the Care and Use of Laboratory
Animals as well as University of California (Los Angeles, CA), Tufts University,
and Mass Eye and Ear Infirmary Institutional Animal Care and Use Committee
Quantitative PCR. SYBR Green-based quantitative RT-PCR was used to assess
were purchased from Qiagen QuantiTect Primer assays library.
Electrophysiological Recordings from Hair Cells. After killing the mice, apical
turns of the organ of Corti were excised from ?10?/?, ?10?/?, ?10?/?, and
?9?/?mice at P8–9 for IHCs and P10–13 for OHCs. These ages were chosen
because they are the times at which maximal ACh-inducible activity is ob-
OHCs were as described in refs. 9, 28, and 47. Recordings were made at room
temperature (22–25°C). Holding potentials were not corrected for liquid
junction potentials (?4mV) or the voltage drop across the uncompensated
series resistance (9–12 M?). All experimental results obtained in IHCs and
OHCs are from two to eight mice of each phenotype.
(IHC region) immunostained with antibody to synaptophysin. (A) ?10?/?mice
exhibit a regular progression of efferent terminals along the ‘‘bottom’’ (mo-
diolar side) of the IHCs (ISB; arrows). These boutons are larger and more
intensely stained than those on the pillar side of the IHC (arrowheads). As a
group of terminals, the efferent system tends to outline the row of IHCs. (B)
?10?/?mice exhibit a disorganized inner spiral bundle (ISB; arrows). The
boutons tend to be large and stain very brightly. Additionally, there is little
the effects of nAChR subunit gene deletion. As previously reported, fewer
pillar side terminals are evident in the ?9?/?ISB, but the modiolar-side
circles indicate the position of the IHCs. (Scale bar is same for A–C, 10 ?m) (D)
Scatter plot of the distance between the nearest neighbor of each terminal in
the field of view. The greater scatter of the ISB terminals in the ?10?/?is
evident in the greater number of terminals separated by ?1.5 ?m. (E) A bar
graph of the mean and SEM further illustrates the change in innervation
between the ?10?/?and ?10?/?mice and lack of effect in ?9?/?mice. Two-
tailed, unpaired t test.***, P ? 6.7 ? 10?7. n.s., not significant.
Whole-mount cochlear turns demonstrating the inner spiral bundle
www.pnas.org?cgi?doi?10.1073?pnas.0708545105Vetter et al.
Cochlear Physiology. Procedures were as described in ref. 34. For recording Download full-text
DPOAEs, animals were anesthetized with ketamine/xylazine. Primary tones f1
and f2(with f2/f1? 1.2 and f2level 10 dB ? f1level) were presented continu-
ously. The ear-canal sound pressure waveform was amplified (?1,000) and
averaged (8 or 25 consecutive waveform traces), and spectrum was computed
by fast Fourier transform. The process was repeated either two or four times,
noise floor (six bins on each side of the DP) were extracted.
ABRs were obtained as described in ref. 7. For OC shock experiments,
animals were anesthetized with urethane (1.20 g/kg i.p.) and surgically pre-
pared as described in refs. 7 and 34. During the OC suppression assay, f2level
was typically set to produce a DPOAE ?10–15 dB above noise floor. To
measure OC effects, repeated measures of baseline DPOAE amplitude were
first obtained (n ? 12), followed by a series of 17 continuous periods in which
DPOAE amplitudes were measured with simultaneous shocks to the OC bun-
dle in the floor of the fourth ventricle.
Immunostaining and Image Processing. Cochleas were fixed in 4% paraformal-
2- to 4-month-old mice were used for all analyses. Efferent terminals were
visualized by using a mouse anti-synaptophysin antibody (MAB5258; Milli-
pore). Sections were slide mounted and cover slipped with SlowFade Gold
(Invitrogen). Samples were examined by using a Leica TCS SP2 AOBS confocal
microscope. For analysis of innervation density in the IHC region, merged Z
manually selected and mapped as points by using the ImageJ plug-in, point-
picker (see http://rsb.info.nih.gov/ij/). The resultant point distribution map
was then imported into R (see www.R-project.org), and nearest-neighbor
distance was calculated by using the SpatStat module (48) of R. A two-tailed
unpaired t test was performed on the nearest-neighbor distances, and a
scattergram plot was generated to illustrate minimum, mean, and maximum
neighbor distances for each synaptic terminal. The mean nearest-neighbor
distance was also graphed as a bar graph with SEMs. All statistical analyses
were performed in Prism (v. 4.0b).
ACKNOWLEDGMENTS. This work was supported by National Institutes of
Health Grants R01 DC6258 (to D.E.V.), R01 DC0188 (to M.C.L.), R21 NS050419
(to D.E.V.), an International Research Scholar Grant from the Howard Hughes
Medical Institute (to A.B.E.), a research grant from Agencia Nacional de
Promocio ´nCientificayTe ´cniaandUniversidaddeBuenosAires,Argentina(to
A.B.E.), and a grant to The Tufts Center for Neuroscience Research (P30
NS047243) supporting the Imaging and Computational Genomics Cores.
1. Hudspeth A (1997) Neuron 19:947–950.
2. Brownell WE, Bader CR, Bertrand D, de Ribaupierre Y (1985) Science 227:194–196.
3. Dallos P (1992) J Neurosci 12:4575–4585.
4. Rasmussen GL (1942) Anat Rec 82:441.
5. Desmedt JE, Lagrutta V (1963) Nature 200:472–474.
6. Walsh E, McGee J, McFadden S, Liberman M (1998) J Neurosci 18:3859–3869.
7. Vetter DE, Liberman MC, Mann J, Barhanin J, Boulter J, Brown MC, Saffiote-Kolman J,
Heinemann SF, Elgoyhen AB (1999) Neuron 23:93–103.
8. Eybalin M (1993) Physiol Rev 73:309–373.
9. Katz E, Elgoyhen A, Go ´mez-Casati M, Knipper M, Vetter D, Fuchs P, Glowatzki E (2004)
J Neurosci 24:7814–7820.
10. Fuchs P, Murrow B (1992) J Neurosci 12:800–809.
11. Blanchet C, Ero ´stegui C, Sugasawa M, Dulon D (1996) J Neurosci 16:2574–2584.
12. Dulon D, Lenoir M (1996) Eur J Neurosci 8:1945–1952.
13. Evans MG (1996) J Physiol (London) 491(Pt 2):563–578.
14. Nenov AP, Norris C, Bobbin RP (1996) Hear Res 101:149–172.
15. Dulon D, Luo L, Zhang C, Ryan A (1998) Eur J Neurosci 10:907–915.
16. Oliver D, Klo ¨cker N, Schuck J, Baukrowitz T, Ruppersberg J, Fakler B (2000) Neuron
17. Glowatzki E, Fuchs P (2000) Science 288:2366–2368.
18. Murugasu E, Russell IJ (1996) J Neurosci 16:325–332.
19. Lustig L, Peng H, Hiel H, Yamamoto T, Fuchs P (2001) Genomics 73:272–283.
20. Luebke AE, Maroni PD, Guth SM, Lysakowski A (2005) J Comp Neurol 492:323–333.
21. Elgoyhen A, Vetter D, Katz E, Rothlin C, Heinemann S, Boulter J (2001) Proc Natl Acad
Sci USA 98:3501–3506.
22. Nie L, Song H, Chen M, Chiamvimonvat N, Beisel K, Yamoah E, Va ´zquez A (2004)
J Neurophysiol 91:1536–1544.
23. Sgard F, Charpantier E, Bertrand S, Walker N, Caput D, Graham D, Bertrand D, Besnard
F (2002) Mol Pharmacol 61:150–159.
24. Maison S, Luebke A, Liberman M, Zuo J (2002) J Neurosci 22:10838–10846.
25. Elgoyhen AB, Johnson DS, Boulter J, Vetter DE, Heinemann S (1994) Cell 79:705–715.
26. Luo L, Bennett T, Jung H, Ryan A (1998) J Comp Neurol 393:320–331.
27. Morley B, Simmons D (2002) Brain Res Dev Brain Res 139:87–96.
28. Go ´mez-Casati M, Fuchs P, Elgoyhen A, Katz E (2005) J Physiol (London) 566:103–118.
29. Hashimoto S, Kimura RS, Takasaka T (1990) Acta Otolaryngol 109:228–234.
30. Sobkowicz HM (1992) in Development of Auditory Systems 2, ed Romand R (Elsevier,
New York) pp, 59–100.
31. Verbitsky M, Rothlin C, Katz E, Elgoyhen A (2000) Neuropharmacology 39:2515–2524.
32. Rothlin CV, Katz E, Verbitsky M, Elgoyhen AB (1999) Mol Pharmacol 55:248–254.
33. Maison S, Adams J, Liberman M (2003) J Comp Neurol 455:406–416.
34. Maison S, Vetter D, Liberman M (2007) J Neurophysiol 97:3269–3278.
36. Weisstaub N, Vetter D, Elgoyhen A, Katz E (2002) Hear Res 167:122–135.
37. Liberman M, Brown M (1986) Hear Res 24:17–36.
38. Folsom RC, Owsley RM (1987) Acta Otolaryngol 103:262–265.
39. Liberman MC (1989) Hear Res 38:47–56.
40. Galambos R (1956) J Neurophysiol 19:424–437.
41. Brown MC, Nuttall AL, Masta RI (1983) Science 222:69–72.
42. Brown MC, Nuttall AL (1984) J Physiol (London) 354:625–646.
43. Mountain, DC (1980) Science 210:71–72.
44. Siegel JH, Kim DO (1982) Hear Res 6:171–182.
45. Franchini LF, Elgoyhen AB (2006) Mol Phylogenet Evol 41:622–635.
46. Brownell WE (1984) Scanning Electron Microsc 1401–1406.
47. Lioudyno M, Hiel H, Kong J, Katz E, Waldman E, Parameshwaran-Iyer S, Glowatzki E,
Fuchs P (2004) J Neurosci 24:11160–11164.
48. Baddeley A, Turner R (2005) J Stat Software 12:1–42.
Vetter et al.
December 18, 2007 ?
vol. 104 ?
no. 51 ?