Multiple quantitative trait loci modify cochlear hair cell degeneration in the Beethoven (Tmc1Bth) mouse model of progressive hearing loss DFNA36.

Abstract

Dominant mutations of transmembrane channel-like gene 1 (TMC1) cause progressive sensorineural hearing loss in humans and Beethoven (Tmc1Bth/+) mice. Here we show that Tmc1Bth/+ mice on a C3HeB/FeJ strain background have selective degeneration of inner hair cells while outer hair cells remain structurally and functionally intact. Inner hair cells primarily function as afferent sensory cells, whereas outer hair cells are electromotile amplifiers of auditory stimuli that can be functionally assessed by distortion product otoacoustic emission (DPOAE) analysis. When C3H-Tmc1Bth/Bth is crossed with either C57BL/6J or DBA/2J wild-type mice, F1 hybrid Tmc1Bth/+ progeny have increased hearing loss associated with increased degeneration of outer hair cells and diminution of DPOAE amplitudes but no difference in degeneration of inner hair cells. We mapped at least one quantitative trait locus (QTL), Tmc1m1, for DPOAE amplitude on chromosome 2 in [(C/B)F1xC]N2-Tmc1Bth/+ backcross progeny, and three other QTL on chromosomes 11 (Tmc1m2), 12 (Tmc1m3), and 5 (Tmc1m4) in [(C/D)F1xC]N2-Tmc1Bth/+ progeny. The polygenic basis of outer hair cell degeneration in Beethoven mice provides a model system for the dissection of common, complex hearing loss phenotypes, such as presbycusis, that involve outer hair cell degeneration in humans.

Full text

Page 1

Copyright ? 2006 by the Genetics Society of America
DOI: 10.1534/genetics.106.057372
Multiple Quantitative Trait Loci Modify Cochlear Hair Cell Degeneration in
the Beethoven (Tmc1Bth) Mouse Model of Progressive Hearing Loss DFNA36
Yoshihiro Noguchi,*,†Kiyoto Kurima,* Tomoko Makishima,*,1Martin Hrabe ´ de Angelis,‡
Helmut Fuchs,‡Gregory Frolenkov,*,2Ken Kitamura†and Andrew J. Griffith*,§,3
*Section on Gene Structure and Function and§Hearing Section, National Institute on Deafness and Other Communication Disorders, National
Institutes of Health, Rockville, Maryland 20850-3320,†Department of Otolaryngology, Tokyo Medical and Dental University
Graduate School, Tokyo 113-8519, Japan and‡GSF Research Center for Environment and Health,
Institute of Experimental Genetics, Neuherberg 85764, Germany
Manuscript received February 17, 2006
Accepted for publication April 26, 2006
ABSTRACT
Dominant mutations of transmembrane channel-like gene 1 (TMC1) cause progressive sensorineural
hearing loss in humans and Beethoven (Tmc1Bth/1) mice. Here we show that Tmc1Bth/1mice on a C3HeB/FeJ
strain background have selective degeneration of inner hair cells while outer hair cells remain structurally
and functionally intact. Inner hair cells primarily function as afferent sensory cells, whereas outer hair cells
are electromotile amplifiers of auditory stimuli that can be functionally assessed by distortion product
otoacoustic emission (DPOAE) analysis. When C3H-Tmc1Bth/Bthis crossed with either C57BL/6J or DBA/2J
wild-type mice, F1hybrid Tmc1Bth/1progeny have increased hearing loss associated with increased degen-
erationofouterhaircellsanddiminutionofDPOAEamplitudesbutnodifferenceindegenerationofinner
hair cells. We mapped at least one quantitative trait locus (QTL), Tmc1m1, for DPOAE amplitude on
chromosome 2 in [(C/B)F13C]N2-Tmc1Bth/1backcross progeny, andthree other QTL onchromosomes 11
(Tmc1m2), 12 (Tmc1m3), and 5 (Tmc1m4) in [(C/D)F13 C]N2-Tmc1Bth/1progeny. The polygenic basis of
outer hair cell degeneration in Beethoven mice provides a model system for the dissection of common,
complex hearing loss phenotypes, such as presbycusis, that involve outerhair cell degeneration in humans.
T
loss(SNHL)inhumans(FriedmanandGriffith2003).
Most autosomal recessive loci are associated with
prelingual severe to profound SNHL, whereas autoso-
mal dominant alleles typically cause postlingual pro-
gressive SNHL (Griffith and Friedman 2002). The
specific causes of SNHL associated with advanced age
(presbycusis) are unknown, but thought to comprise a
complex combination of genetic and environmental
factors (Schultz et al. 2005). Mutant mice are impor-
tant tools for identifying these factors, their function in
the auditory system, and the pathogenesis of hearing
loss (Haider et al. 2002). For example, studies of poly-
genic age-related hearing loss in inbred mouse strains
(Noben-Trauthet al. 2003) facilitated the recent iden-
tificationofa genetic modifier ofhearing loss inhumans
(Schultz et al. 2005), demonstrating the applicability
of mouse models to the dissection of complex hearing
loss traits in humans.
HERE are .100 loci at which mutations cause
monogenic nonsyndromic sensorineural hearing
Dominant and recessive mutations of transmem-
brane channel-like gene 1 (TMC1) cause nonsyndromic
SNHLattheDFNA36andDFNB7/B11loci,respectively
(Kurima et al. 2002, 2003). TMC1 encodes a polytopic
transmembrane protein of unknown function that is
expressed in cochlear hair cells (Kurima et al. 2002;
Vreugde et al. 2002). In mice, dominant and recessive
mutations of Tmc1 cause SNHL in the Beethoven (Bth)
and deafness (dn) mouse lines, respectively (Kurima
et al. 2002; Vreugde et al. 2002). Whereas DFNB7/B11
and dn homozygotes have severe to profound congenital
hearing loss, heterozygous carriers of Bth and DFNA36
mutations have delayed-onset, progressive SNHL. SNHL
in Beethoven and deafness mice is associated with rapid
degeneration of cochlear hair cells (Bock and Steel
1983; Vreugde et al. 2002), indicating that Tmc1 is
required for normal hair cell function or survival.
The mammalian cochlea contains two types of hair
cells distinguished by their location, morphology, and
function (Frolenkov et al. 2004). In general terms,
inner hair cells (IHCs) have primarily afferent innerva-
tion and function as sensory cells transducing and
transmitting auditory signals to the central nervous
system. Outer hair cells (OHCs) receive principally
efferent innervation and have a distinctive electromo-
tile property postulated to underlie active biomechan-
ical amplification of auditory stimuli. As a result of this
1Present address: Department of Otolaryngology, University of Texas
Medical Branch, Galveston, TX 77555-0521.
2Present address: Department of Physiology, University of Kentucky,
Lexington, KY 40536-0298.
3Corresponding author: National Institutes of Health, 5 Research Court,
Rockville, MD 20850-3320. E-mail: griffita@nidcd.nih.gov
Genetics 173: 2111–2119 (August 2006)

Page 2

active amplification, OHCs generate sounds known as
otoacoustic emissions (OAEs) that can be measured
noninvasively in living humans and mice with a sensitive
microphone in the external auditory canal. In contrast,
evaluation of hearing levels by auditory brainstem re-
sponse(ABR)thresholdanalysisisanoverallmeasureof
peripheral auditory function, including both inner
and outer hair cell function. Primary damage or sec-
ondary degeneration of OHCs characterizes many com-
mon hearing loss phenotypes, such as presbycusis, in
humans. Since genetic modifiers are excellent candi-
dates for etiologic determinants of these complex traits
(Friedman et al. 2000; Schultz et al. 2005), including
presbycusis, we sought to identify modifiers of hair cell
degeneration in Beethoven mice.
MATERIALS AND METHODS
Experimental animals: All protocols were approved by the
National Institute of Neurological Disorders and Stroke/
National Institute on Deafness and Other Communication
Disorders Animal Care and Use Committee. Isogenic hetero-
zygous Beethoven (Tmc1Bth/1) mice on C3HeB/FeJ (C3H)
(Vreugde et al. 2002) were intercrossed to generate homozy-
gous (C3H-Tmc1Bth/Bth), heterozygous (C3H-Tmc1Bth/1), and
wild-type (C3H-Tmc11/1) animals. C3H-Tmc1Bth/Bthhomozy-
gotes were crossed with wild-type C57BL/6J or DBA/2J mice
to generate (C/B)F1-Tmc1Bth/1and (C/D)F1-Tmc1Bth/1hybrids,
respectively. F1hybrids were backcrossed to wild-type parental
strains to generate [(C/B)F13 C]N2-Tmc1Bth/1and [(C/D)F13
C]N2-Tmc1Bth/1Beethoven progeny.
Tmc1 genotype analysis: Genomic DNA was prepared from
tail clip biopsies by a phenol/chloroform extraction pro-
cedure. A 213-bp fragment of Tmc1 exon 13 was amplified
with forward (59-TAT TAA AGGGAC CGC TCT GAA AAC-39)
and reverse (59-ATC CAT CAA GGC GAG AAT GAA TAC-39)
primers in a 20-ml volume containing 50 ng DNA, 20 pmol of
each primer, 200 mmol/liter of each dNTP, 1.5 mmol/liter
MgCl2, and 1.6 units Taq DNA polymerase. Amplification
conditions comprised an initial 2-min denaturation at 95?,
followed by 30 step-cycles of 30 sec at 95?, 30 sec at 57?, and 45
secat72?,withafinalelongationof5minat72?.PCRproducts
were directly sequenced on an automated sequencer (ABI-
PRISM, model 3700; Applied Biosystems, Foster City, CA).
ABR and distortion product otoacoustic emissions ana-
lyses: ABR thresholds were measured as described (Szymko-
Bennett et al. 2003) with some modifications: we used
alternating polarity click and tone-burst stimuli of 47-msec
and 5-msec duration, respectively. The number of stimulus
presentations was varied from 128 to 1024 depending on
signal-to-noise ratio, and suprathreshold stimulus intensities
wereinitiallydecreasedin10-dBstepsfollowedby5-dBstepsat
lower intensities to determine the response threshold. When
no waveform was detectable at the highest stimulus level of
90 dB sound pressure level (SPL), the threshold was consid-
ered to be 95 dB SPL for subsequent analyses.
Distortion product otoacoustic emissions (DPOAEs) were
recorded with an acoustic probe (ER-10C; Etymotic Research,
Elk Grove Village, IL) using DP2000 DPOAE measurement
system version 3.0 (Starkey Laboratory, Eden Prairie, MN).
Two primary tones with frequency ratio f2/f1 ¼ 1.2 were
presented at intensity levels L1¼ 65 dB SPL and L2¼ 55 dB
SPL. f2was varied in one-sixth-octave steps from ?4 to 16 kHz.
DP-gramscomprised2f1-f2DPOAEamplitudesasafunctionof
f2. Due to the limited frequency response of the acoustical
transducers of the ER-10C probe, the software was able to
correct stimulus intensities only at frequencies ,16 kHz.
Therefore, reliable DPOAE measurements were possible only
at f2frequencies ,16 kHz. Table 1 shows the numbers of mice
tested by ABR and DPOAE analyses.
Histologic analysis: Organ of Corti specimens were dis-
sected, fixed in 4% paraformaldehyde, and stained with Alexa
Fluor 568 phalloidin (Molecular Probes, Eugene, OR) to
visualize hair cells and stereocilia as described (Belyantseva
etal.2005).Innerandouterhaircellswerecountedinacentral
segment of eachof threeregions at10–20, 40–50, and70–80%
of the total cochlear duct distance from the apex, approxi-
mately corresponding to frequency ranges of 7–8, 15–18, and
36–42 kHz (Viberg and Canlon 2004), respectively. Each
segment contained a sum total of ?80 hair cell positions/row
with an intact, degenerated, or lost hair cell. Hair cells were
counted in at least six mutant or four control ears for each
genotype and developmental time point and were considered
to be degenerated if the cell soma or stereocilia were absent.
Table 2 shows the numbers of mice examined histologically.
Short tandem repeat genotype analysis: Short tandem
repeat (STR) markers (listed in Table 3) spanning the mouse
genome were selected at ?20-cM intervals. The Jackson
Laboratory Mouse Genome Informatics website (http:/ /www.
informatics.jax.org/) was the source of genetic map positions
of STR markers. We analyzed additional markers (listed in
Table 4) at loci with significant or suggestive linkage. PCR
amplification conditions comprised an initial denaturation
stepof95?for 2min,33step-cycles of95?for 30sec,51–60?for
30 sec, and 72? for 45 sec, followed by a final elongation at 72?
for 5 min. Amplification products were separated by 3% aga-
rose (NuSieve 3:1; Cambrex Bio Science Rockland, Rockland,
ME) gel electrophoresis and visualized by ethidium bromide
staining and ultraviolet illumination.
QTL association analysis: The 2f1-f2DPOAE amplitudes of
P56 N2mice were determined at three frequencies (f2¼
12,609,14,156,and15,891Hz)inbothears.Thesefrequencies
correspond to a cochlear duct region ?40% of the total duct
distance from the apex (Viberg and Canlon 2004). The
average of the six DPOAE values was analyzed for association
with STR marker genotypes on a Macintosh computer with
QTX software downloaded from http:/ /www.mapmanager.
org/mmQTX.html (Manly et al. 2001). The significance
threshold of likelihood ratio statistic (LRS) ¼ 12.4 for interval
mapping was estimated by the Quick Test application of QTX
(Manly et al. 2001). Conventional base-10 LOD scores were
calculated by dividing the LRS by 4.61 (Manly et al. 2001).
Statistical analysis: The significance of differences in ABR
thresholds and DPOAE amplitudes was calculated by Mann–
WhitneyUtestingwithStatViewversion4.51software (Abacus
Concepts, Berkeley, CA) on a Macintosh computer.
RESULTS
Hearing thresholds: The mouse auditory system is
functionally mature by 3 weeks of age (Ehret 1977). At
3weeksofage,homozygousC3H-Tmc1Bth/Bthmicehadno
detectable ABR to click stimuli, whereas C3H-Tmc1Bth/1
mice had slightly elevated thresholds in comparison
to C3H-Tmc11/1littermates (Figure 1A). C3H-Tmc1Bth/1
ABR thresholds for both click and tone-burst stim-
uli steadily increased to reach undetectable levels by
24 weeks of age, whereas C3H-Tmc11/1thresholds re-
mainedstable (Figure 1;tone-burstresponsethresholds
not shown). These results are consistent with reported
2112Y. Noguchi et al.

Page 3

compound action potential threshold recordings in-
dicating the semidominant nature of Tmc1Bth(Vreugde
et al. 2002).
To determine if strain background affects phenotypic
expression of Tmc1Bth, we crossed C3H-Tmc1Bth/Bthwith
either DBA/2J or C57BL/6J mice and evaluated audi-
tory function of F1hybrid progeny from 3 weeks to 20
weeks of age (Figure 1B). In comparison to isogenic
C3H-Tmc1Bth/1
mice, (C/B)F1-Tmc1Bth/1
thresholds were modestly but significantly higher (P ¼
0.0016, 4 weeks; P , 0.0001, 8 weeks; P ¼ 0.0001, 12
weeks; P ¼ 0.0007, 16 weeks; P , 0.0001, 20 weeks). (C/
D)F1-Tmc1Bth/1click thresholds were intermediate be-
tween those of C3H-Tmc1Bth/1and (C/B)F1-Tmc1Bth/1
mice from 4 to 12 weeks of age, but still significantly
higher than C3H-Tmc1Bth/1thresholds at all but one of
these developmental time points: P ¼ 0.0055, 4 weeks;
P ¼ 0.0718, 8 weeks; P ¼ 0.02, 12 weeks; P ¼ 0.0003, 16
weeks; P , 0.0001, 20 weeks. (C/D)F1-Tmc1Bth/1click
thresholdsweresignificantlylowerthan(C/B)F1-Tmc1Bth/1
thresholds only at the 8- and 12-week time points: P ¼
0.2134, 4 weeks; P ¼ 0.0009, 8 weeks; P ¼ 0.0019, 12
weeks; P ¼ 0.9625, 16 weeks; P . 0.9999, 20 weeks.
Figure 1B shows that these strain background effects on
ABR thresholds were largely specific for Tmc1Bth/1mice
since similar differences were not observed with C3H-
Tmc11/1mice in comparison to (C/B)F1-Tmc11/1(P ¼
0.4824, 4 weeks; P ¼ 0.6222, 8 weeks; P ¼ 0.2908, 24
weeks) or (C/D)F1-Tmc11/1mice (P ¼ 0.4606, 4 weeks;
P ¼ 0.6222, 8 weeks; P ¼ 0.0361, 24 weeks).
Cochlear hair cell degeneration: We evaluated IHC
and OHC degeneration in basal, middle, and apical
regions of the cochlear ducts of isogenic Beethoven
mice. At 4 weeks of age, C3H-Tmc1Bth/Bthcochleae had
nearly complete IHC degeneration in all three regions,
whereas OHCs showed a tonotopic gradient of de-
generation with a few percent degeneration in the
apical duct, 30% in the middle turn, and 70% in the
basal turn (Figure 2). In contrast, C3H-Tmc11/1co-
chleae had no degeneration of IHCs or OHCs at 20
weeks of age. C3H-Tmc1Bth/1hair cells showed an inter-
ABRclick
mediate phenotype. At all developmental time points,
C3H-Tmc1Bth/1OHC degeneration was limited to the
basal region, progressing from 20% degeneration at
4 weeks of age to 40% at 20 weeks of age. C3H-Tmc1Bth/1
IHC degeneration was partial (70%) and mainly re-
stricted to the basal region at 4 weeks of age, but
increased in a base-to-apex direction with increasing
age. By 20 weeks of age, IHC degeneration was essen-
tially complete in the basal and middle regions, and
40% complete in the apical region.
To identify strain-dependent effects on hair cell
degeneration, we compared isogenic C3H-Tmc1Bth/1
mice with (C/D)F1-Tmc1Bth/1and (C/B)F1-Tmc1Bth/1hy-
brids at 8 weeks of age. There were no detectable dif-
ferences in the degree or location of IHC degeneration
(Figure 2). However, there was greater OHC degener-
ationinthehybridgroupsincomparisontotheisogenic
mutants: 100 vs. 35% degeneration in the basal region
and ?40 vs. 0% in the middle region, respectively.
There were no significant differences in OHC degener-
ationbetweenthetwohybridgroups:P¼0.3068,apical;
TABLE 1
Number of mice tested in ABR and DPOAE analyses
Age (wk)
Strain-genotype34568 1216 20 24
C3H-Tmc11/1
(C/B)F1-Tmc11/1
(C/D)F1-Tmc11/1
C3H-Tmc1Bth/Bth
C3H-Tmc1Bth/1
(C/B)F1-Tmc1Bth/1
(C/D)F1-Tmc1Bth/1
0/21
0/6
10/12
6/6
5/5
5/7
15/15
12/12
15/15
0/20/2 9/14
11/11
5/5
5/5
11/11
11/11
10/10
5/5
0/6
0/15
0/12
4/4
15/15
27/27
14/14
0/15
0/10
0/15
0/14
0/10
12/12
15/20
14/14
11/11
17/17
11/11
11/11
17/17
11/11
10/10
Number of mice tested in both ABR and DPOAE analyses are shown on the left and total number of mice tested by ABR analysis
are shown on the right.
TABLE 2
Number of mice examined histologically
Distance from apex
Strain-genotypeAge (wk)10–20%40–50%70–80%
C3H-Tmc1Bth/1
C3H-Tmc1Bth/Bth
C3H-Tmc11/1
C3H-Tmc1Bth/1
C3H-Tmc1Bth/Bth
C3H-Tmc11/1
C3H-Tmc1Bth/1
C3H-Tmc11/1
C3H-Tmc1Bth/1
C3H-Tmc11/1
(C/B)F1-Tmc1Bth/1
(C/B)F1-Tmc11/1
(C/D)F1-Tmc1Bth/1
(C/D)F1-Tmc11/1
4
4
4
8
8
8
11
8
8
10
10
8
15
5
8
4
10
7
12
4
7
8
7
9
7
7
7
6
4
6
7
4
8
4
7
4
7
6
12
12
20
20
8
8
8
8
13
6
8
4
8
8
11
4
11
4
Modifiers of Hair Cell Degeneration2113

Page 4

P ¼ 0.2477, middle; P ¼ 0.9639, basal. Whereas OHC
degeneration was similar among the three rows at any
particular location in both hybrid groups, the inner row
was most affected and theouter row was least affected in
C3H-Tmc1Bth/1cochleae (not shown).
Outer hair cell function: We measured DPOAEs to
determine the effect of strain background on OHC
function in Beethoven mice. Four- and 12-week-old
C3H-Tmc1Bth/Bthmice had DPOAE amplitudes that were
not different from the noise floor of ?10 to ?20 dB SPL
(Figure 3, A and C) in comparison to robust DPOAEs of
.30 dB SPL that increased slightly in amplitude from 4
through 32 weeks of age in Tmc11/1mice (Figure 3, A
and D). C3H-Tmc1Bth/1DPOAEs had a similar configu-
ration and agedependence (Figure 3,Aand B)but with
?5 dB lower amplitudes than wild-type DPOAEs at most
frequencies. At 4 weeks of age, these differences were
significant (P , 0.05) for all f2 frequencies above
?5 kHz. There were no differences at low f2frequencies
since even wild-type DPOAEs are typically undetectable
above acoustic noise for f2# 5 kHz (Figure 3A).
In comparison to C3H-Tmc1Bth/1mice, DPOAE ampli-
tudes were significantly decreased at the higher mea-
sured f2frequencies in (C/B)F1-Tmc1Bth/1(P , 0.0001
for f2¼ 11,203, 12,609, 14,156, and 15,891 Hz) and (C/
D)F1-Tmc1Bth/1mice (P , 0.0001 for f2¼ 8906, 9984,
11,203, 12,609, 14,156, and 15,891 Hz) (Figure 4A).
DPOAE amplitudes were significantly (P , 0.01) higher
in (C/B)F1-Tmc1Bth/1mice than in (C/D)F1-Tmc1Bth/1mice
only at lower f2frequencies (4453, 5016, 5625, 6281, 7078,
7922, and 8906 Hz). The decrease in these DPOAE am-
plitudes in the hybrid mutant groups occurred between
4and8weeksofage(Figure4,CandD).Corresponding
DPOAE amplitudes in F1 hybrid wild-type controls
were identical or slightly increased in comparison with
isogenic wild-type controls (Figure 4B), indicating the
effect of background strain on DPOAE amplitudes is
specific for heterozygous carriers of Tmc1Bth/1.
TABLE 3
STR markers used in genomewide QTL association analysis
MarkerMap position (cM) MarkerMap position (cM)Marker Map position (cM)
D1Mit212
D1Mit132
D1Mit442
D1Mit15
D1Mit223
21
43.1
63.1
87.9
106.3
D7Mit56
D7Mit247
D7Mit318
D7Mit 105
2.5
16
37
65
D14Mit140
D14Mit63a
D14Mit262b
D14Mit92
14.5
22.5
28.4
45
D8Mit190b
D8Mit301a
D8Mit162
D8Mit215
D8Mit280
21
32
37
59
72
D15Mit102
D15Mit128
D15Mit105
D15Mit246
7
D2Mit464a
D2Mit369b
D2Mit323
D2Mit92
D2Mit444
D2Mit307
D2Mit59
9.5
27.3
31.7
41.4
65.8
74.9
86
26
48
60.1
D16Mit38
D16Mit30
D16Mit76b
D16Mit50a
28.5
36
43
53.5
D9Mit224
D9Mit208
D9Mit308
D9Mit137
17
35
49
66 D3Mit221
D3Mit333
D3Mit77
D3Mit292
2.4
22
49.7
72.9
D17Mit164
D17Mit51
D17Mit238
D17Mit155
4
D10Mit80
D10Mit106
D10Mit115
D10Mit180
4 23
35
55
17
38.4
64 D4Mit236
D4Mit53
D4Mit26
D4Mit134
D4Mit190
12.1
19.8
42.5
62.3
79
D18Mit94
D18Mit105
D18Mit40
D18Mit80b
D18Mit128a
17
25
37
50
55
D11Mit2
D11Mit313
D11Mit320
D11Mit288
D11Mit253
2.4
28
43
55
71D5Mit349
D5Mit394
D5Mit205
D5Mit23
D5Mit292
8
34
45
54
80
D19Mit68
D19Mit41
D19Mit11
D19Mit91
6
D12Mit270
D12Mit52
D12Mit20
13
32
58
16
41
47
D6Mit268
D6Mit146
D6Mit366
15.6
38.5
50.5
D13Mit266
D13Mit64
D13Mit226
16
30
59
aUsed only in [(C/D)F13 C]N2-Tmc1Bth/1mice.
bUsed only in [(C/B)F13 C]N2-Tmc1Bth/1mice.
2114 Y. Noguchi et al.

Page 5

Modifier loci: The different phenotypes of isogenic
and hybrid Tmc1Bth/1mice suggested the existence of
oneormoremodifierlociwithstrain-specificallelesthat
differentially affect OHC degeneration caused by Tmc1Bth.
It also indicated that, for at least one modifier locus,
DBA/2J and C57BL/6J alleles are dominant or semi-
dominantin trans with C3HeJ/FeB alleles. Wetherefore
backcrossed F1hybrids to their parental strains and
analyzed DPOAEs in their N2offspring at 8 weeks of
age. We initially observed a multimodal distribution
of DPOAE amplitudes in [(C/D)F13 C]N2-Tmc1Bth/1
and [(C/B)F13 C]N2-Tmc1Bth/1mice that suggested the
existence of dominant or semidominant DBA/2J and
C57BL/6J allelesat oneortwo modifierloci(not shown).
Total or near-total loss of DPOAEs in [(C/D)F13 D]N2-
Tmc1Bth/1and [(C/B)F13 B]N2-Tmc1Bth/1progeny were
consistent with a semidominant model (not shown).
To map loci modifying OHC degeneration in Tmc1Bth/1
mice, we measured DPOAEs of 144 [(C/B)F13 C]N2-
Tmc1Bth/1and 180 [(C/D)F13 C]N2-Tmc1Bth/1backcross
progeny (Figure 5). The continuous distribution of
DP-grams was not consistent with a simple model of
mono- or digenic inheritance of modifier alleles, so we
analyzed a three-frequency DPOAE amplitude average
as a quantitative trait (see materials and methods).
The heritability of this trait was 0.82 in [(C/B)F13
C]N2-Tmc1Bth/1mice and 0.86 in [(C/D)F1 3 C]N2-
Tmc1Bth/1mice. The subtotal heritability is consistent
with the degree of variation of DPOAE amplitudes in F1
hybrid mice (Figure 4).
Short tandem repeat genotypes were determined in
15 mice with the highest DPOAE amplitude averages
and 15 mice with the lowest averages among the N2
progeny of each backcross. These selective genotype
analyses identified a potential locus (Tmc1m1) on chro-
mosome 2 in the C57BL/6J backcross and a locus
(Tmc1m2) on chromosome 11 in the DBA/2J back-
cross (not shown). Genotypes of remaining backcross
progeny at these loci revealed only weak associations
with DPOAE amplitude, indicating that additional
loci probably contribute to modification of OHC loss
in Tmc1Bth/1mice. Due to the selection for N2mice car-
rying Tmc1Bth, which is located on chromosome 19, STR
markers on chromosome 19 inherited from the F1
parent were nearly always C3HeB/FeJ alleles in linkage
disequilibrium with Tmc1.
To detect additional modifier loci, we performed a
genomewide QTL association analysis on a total of 144
[(C/B)F13 C]N2-Tmc1Bth/1and 125 [(C/D)F13 C]N2-
Tmc1Bth/1backcrossprogeny(Table5,Figure6).Interval
mapping confirmed the strong association on chromo-
some 11 in the DBA/2J backcross progeny (Figure 6B)
and weaker but significant association with chromo-
some2intheC57BL/6Jbackcrossprogeny(Figure6A).
Although the appearance of two peaks might indicate
that more than one locus may exist in each interval,
genotype analysis of additional markers (listed in
Table 4) and additional animals (55 [(C/D)F13 C]N2-
Tmc1Bth/1mice) could not significantly demonstrate
the existence of two distinct loci within either broad
interval. Tmc1m1 and Tmc1m2 accounted for 16 and
19% of the observed trait variation in the respective N2
backcross progeny (Table 5). We also detected sig-
nificant associations of reduced DPOAE amplitudes
with DBA/2J alleles of markers on chromosomes 12
(Tmc1m3; Figure 6C) and 5 (Tmc1m4; Figure 6D) that
TABLE 4
QTX statistics for QTL affecting DPOAE amplitude
Symbol
Tmc1m1Tmc1m2Tmc1m3 Tmc1m4
——
Chromosome
Position (cM)
LOD score
Variance (%)
Protective allelea
Susceptibility allelea
2 11
55
8.5
19
C
D
12
3
4.3
10
C
D
5 15
60
17
2546
5.8
16
C
B
61
3.6
9
C
D
2.8
9
B
C
2.7
8
B
C
Dashes indicate loci that did not reach statistical significance.
aStrain alleles: C, C3HeB/FeJ; B, C57BL/6J; D, DBA/2J.
Figure 1.—Mean click-evoked ABR thresholds for right
ears of Beethoven mice. (A) ABR thresholds for isogenic
Tmc11/1(C3H-1/1), Tmc1Bth/1(C3H-Bth/1), and Tmc1Bth/Bth
(C3H-Bth/Bth) mice. (B) ABR thresholds for (C/B)F1-
Tmc1Bth/1(C/B-Bth/1), (C/D)F1-Tmc1Bth/1(C/D-Bth/1),
(C/B)F1-Tmc11/1(C/B-1/1), and (C/D)F1-Tmc11/1(C/D-
1/1) hybrid mice. ABR thresholds for isogenic C3H-
Tmc11/1and C3H-Tmc1Bth/1mice are shown as in A. When
responses were undetectable with a 95 dB SPL stimulus, the
threshold was calculated and presented as 95 dB SPL. Vertical
bars indicate standard deviations.
Modifiers of Hair Cell Degeneration 2115