The transmembrane inner ear (Tmie) protein is essential for normal hearing and balance in the zebrafish Michelle R. Gleasona, Aaron Nagiela, Sophie Jametb,1, Maria Vologodskaiaa, Herna´n Lo´pez-Schiera,b,2, and A. J. Hudspetha,2 aHoward
Hughes Medical Institute and Laboratory of Sensory Neuroscience, The Rockefeller University, New York, NY 10065; and bLaboratory of Sensory Cell Biology and Organogenesis, Centre de Regulacio´ Geno`mica-Barcelona Biomedical Research Park (PRBB), 08003 Barcelona, Spain Contributed by A. J. Hudspeth, October 7, 2009 (sent for review September 3, 2009)
auditory system 兩 hair cell 兩 lateral line 兩 mechanoelectrical transduction 兩 vestibular system
T
he vertebrate inner ear is a complex organ that houses the delicate mechanoreceptors for hearing and balance known as hair cells. Situated at the apical surface of each hair cell is an array of height-ordered stereocilia called the hair bundle. This mechanically sensitive organelle responds in a directiondependent manner to displacements caused by sound and acceleration (1). Deflection of the hair bundle toward its tall edge increases the open probability of transduction channels residing at the stereociliary tips, allowing a depolarizing inward current of cations from the surrounding endolymph. Bundle movement in the opposite direction has the opposite effect, hyperpolarizing the hair cell. The conversion of acoustic and accelerational stimuli into electrical signals that are transmitted to the auditory nerve and subsequently to the brain is known as mechanoelectrical transduction. Although we have a basic understanding of the biophysical and electrophysiological events that initiate hearing, we know much less about the molecules involved. The difficulty in identifying these components stems from the paucity of sensory tissue in the inner ear, which frustrates biochemical purification and traditional molecular-biological assays. Genetic investigation has therefore largely replaced these approaches as the preferred strategy for identifying proteins important in hearing (2). The zebrafish has proven useful for this purpose (3). The internal ear of the zebrafish, which is anatomically and functionally similar to those of other vertebrates (4), undergoes rapid development and is readily accessible for observation and manipulation owing to its optical transparency. The zebrafish possesses an additional feature useful for hair-cell investigations, the lateral-line system. This apparatus, which comprises a series of hair-cell clusters termed neuromasts distributed over the body surface, is used for the detection of water movements, for www.pnas.org兾cgi兾doi兾10.1073兾pnas.0911632106
schooling, and for the detection of prey and predators (5). Genetic investigations of zebrafish with abnormal hearing and vestibular dysfunction have identified several proteins whose absence impairs the development or operation of hair cellcontaining organs (6–9). In the present study, we report the results of an investigation of a mutant with extreme deficits of both auditory and vestibular transduction. Results Characteristics of Mutant Larvae. While examining Tg(Hsp70:eGFP)
larvae (10), we noticed that approximately one-quarter failed to startle in response to acoustic stimuli. Because the animals reacted to gentle touch with C-start avoidance behavior (11), the lack of acoustic responsiveness was not the result of an inability to initiate movement. Although some mutants failed to inflate their swim bladders, others did so but floated on their sides at the water’s surface, apparently incapable of orienting themselves upright when not swimming. Motility was uncoordinated, often with rapid circling and spiraling motions. Mutants could nevertheless survive to adulthood as homozygotes with persistent vestibular dysfunction. Hair-Cell Labeling. To determine whether the defect occurred at
the level of the sensory hair cells, we immersed living larvae in a solution of either 4-Di-2-ASP (4-(4-(diethylamino)styryl)-Nmethylpyridinium iodide) or FM4–64 (N-(3-triethylammoniumpropyl)-4-(6-(4-(diethylamino)phenyl)hexatrienyl)pyridinium dibromide). These cationic fluorophores are thought to label hair cells by traversing the mechanosensitive channels, thus providing an index of mechanotransduction (12, 13). When mutant fish were subjected to this assay, we observed no fluorophore incorporation in the neuromasts of the lateral-line organ despite the unquestionable presence of hair cells, which were labeled by Tg(pou4f3:gap43-mGFP) (14) (Fig. 1 A and B). The hair cells in the inner ears of mutants also failed to incorporate fluorophore that was pressure-injected into the otic cavity, whereas WT hair cells were robustly labeled by this technique (Fig. 1 C and D). Microphonic Recordings. To document the defect in hair-cell
function by electrophysiological means, we recorded the microphonic potentials evoked by vibration of the inner ear with a glass probe driven by a piezoelectric stimulator. Microphonic Author contributions: M.R.G., H.L.-S., and A.J.H. designed research; M.R.G., A.N., S.J., M.V., and H.L.-S. performed research; M.R.G. and H.L.-S. analyzed data; and M.R.G., A.N., S.J., M.V., H.L.-S., and A.J.H. wrote the paper. The authors declare no conflict of interest. Freely available online through the PNAS open access option. 1Present
address: UMR 7102 CNRS, Universite´ Pierre et Marie Curie, 9, Quai St Bernard, 75005 Paris, France.
2To
whom correspondence may be addressed. E-mail:
[email protected] or hudspaj@ rockefeller.edu.
PNAS 兩 December 15, 2009 兩 vol. 106 兩 no. 50 兩 21347–21352
NEUROSCIENCE
Little is known about the proteins that mediate mechanoelectrical transduction, the process by which acoustic and accelerational stimuli are transformed by hair cells of the inner ear into electrical signals. In our search for molecules involved in mechanotransduction, we discovered a line of deaf and uncoordinated zebrafish with defective hair-cell function. The hair cells of mutant larvae fail to incorporate fluorophores that normally traverse the transduction channels and their ears lack microphonic potentials in response to vibratory stimuli. Hair cells in the posterior lateral lines of mutants contain numerous lysosomes and have short, disordered hair bundles. Their stereocilia lack two components of the transduction apparatus, tip links and insertional plaques. Positional cloning revealed an early frameshift mutation in tmie, the zebrafish ortholog of the mammalian gene transmembrane inner ear. The mutant line therefore affords us an opportunity to investigate the role of the corresponding protein in mechanoelectrical transduction.
A
A
B
B
C
D
B′
C′
D′
PO AO
A′ PO AO
C
D
OV
F
E F
G G
PC AC
OV
E′
F′
G′
MC
PC
H
I
J
K
H′
I′
J′
K′
AC MC
E
Wild-type Mutant
5 ms
Stimulus Fig. 1. Defective hair-cell function. (A) In a WT (Upper) and a mutant (Lower) larva at 6 dpf, arrowheads indicate neuromasts labeled by Tg(pou4f3:gap43mGFP) (green). (B) Upon exposure to FM4 – 64 (red), the WT specimen (Upper) shows robust incorporation of the fluorophore into hair cells but the mutant (Lower) shows none. (C) In the inner ears of WT (Upper) and mutant (Lower) animals at 5 dpf, Tg(pou4f3:gap43-mGFP) (green) marks the anterior (AC), medial (MC), and posterior cristae (PC). (D) Pressure-injection of FM4 – 64 (red) into the otic cavities labels hair cells in the cristae of WT (Upper) but not mutant (Lower) fish. (E) Averaged extracellular recordings of microphonic potentials measured at 5–7 dpf from the inner ears of eight WT larvae show a clear response at twice the stimulus frequency, whereas the record from eight mutant larvae shows only a stimulus artifact. (Scale bars, 200 m in A and B; 20 m in C and D.)
potentials are an extracellular manifestation of the transduction currents carried by hair cells in the inner ear (15). Although WT fish responded with robust microphonic potentials, mutants lacked the typical responses at a frequency twice that of the stimulus (Fig. 1E). Application of gentamicin, a reversible blocker of mechanotransduction channels, completely abolished the response in WT fish; the drug had no effect on the signal recorded from mutant larvae, confirming that it was a stimulus artifact. Phenotype. The mutants lack gross morphological or develop-
mental defects of the body and inner ears, displaying two 21348 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0911632106
Fig. 2. Absence of gross morphological and developmental defects. In each panel, a WT larva (Upper) is contrasted with a mutant (Lower). (A, A⬘) Mutants display a normal body shape and length at 4 dpf. (B, B⬘) In a mutant, the semicircular canals develop normally and the anterior (AO) and posterior otoliths (PO) are normal. (C, C⬘) Labeling of hair cells by Tg(pou4f3:gap43mGFP) (green) reveals no obvious differences in their number, shape, or localization in the anterior macula. (D, D⬘) In lateral-line neuromasts at 3 dpf, the hair bundles of mutants display normal planar cell polarity. The kinocilium of each hair bundle is marked by a black notch denoting the absence of actin staining by phalloidin. At 22 hpf (E, E⬘), 2 dpf (F, F⬘), and 3 dpf (G, G⬘), mutants display normal migration of the lateral-line primordium and formation of neuromasts, here labeled by Tg(cldnB:lynGFP) (green). OV, otic vesicle; arrowheads indicate the leading edges of the primordia. (H and I) Supporting cells labeled by Tg(cldnB:lynGFP) (green) are normally arranged in mutants in the anterior (H, H⬘) and posterior lateral lines (I, I⬘) of 5-dpf larvae. (J, J⬘) Fifty hours after treatment with Cu2⫹, a 7-dpf mutant larva shows normal regeneration of hair cells labeled by Et(krt4:GFP)sqet4 (green). (K, K⬘) Phalloidin staining shows that regenerated hair cells display planar cell polarity in a mutant. (Scale bars, 500 m in A, F, and G; 50 m in B; 5 m in C; 2 m in D and K; 100 m in E; 10 m in H–J.)
normal-sized otoliths correctly associated with sensory maculae and three cristae of the semicircular canals (Fig. 2 A and B). The hair cells themselves appear similar to those in WT siblings in shape, size, and position (Fig. 2C). Actin staining also showed normal planar cell polarity in mutants (Fig. 2D). To rule out other possible causes of the mutant phenotype, we crossed mutant carriers to Tg(cldnB:lynGFP) fish (16) and observed the lateral-line system. The pattern and timing of the migration of the posterior lateral-line primordium, the establishment of the lateral-line neuromasts, and the organization of their supporting cells all appeared normal (Fig. 2 E–I). Using the Et(krt4:GFP)sqet4 line (17) to visualize recovery after Cu2⫹Gleason et al.
Z57770
fk73d01
ccdc13
1.1
0.1
0.04
fk73d01
ccdc13
tmie
pthr1
Z44904
5.5 cM
0.06
tmie
pthr1
CH211-163F10 43
B
Wild-type
GCGATGGAGGGGCCTG A M E G P
41
31
Homozygous mutant GCGATG A AGGGCCTG A M K G L
C
D 5’ E1
E2 E3 E4a
E4
3’
kb Heterozygous mutant GCGATG - AGGG - C -G A M G -
COOH C M E NH2
E
MRRGRRRGKMAMEGPGNQPPSLVPLLMSVTAVLI MRRGRRRGKMAMEGPGNQPPSLVPLLMSVTAVLI ---------MAGRQHGSGR------LWALGGAAL ---------MAG-WPGAGP------LCVLGGAAL
34 34 19 18
SHHFFNAAAQIPDPELLPTDPPKKPDPVTSETVVFWGLRLWQVVGIFSMF SHHFFNAAAQIPDPELLPTDPPKKPDPVTSETVVFWGLRLWQVVGIFSMF GACLAGVATQLVEPSTAPPKP--KPPPLTKETVVFWDMRLWHVVGIFSLF GVCLAGVAGQLVEPSTAPPKP--KPPPLTKETVVFWDMRLWHVVGIFSLF
84 84 67 66
ILAIIITLCCIFKCRIPRTKKEIEARHAQRLAAKKYANTLETVPPLNELT ILAIIITLCCIFKCRIPRTKKEIEARHAQRLAAKKYANTLETVPPLNELT VLSIIITLCCVFNCRVPRTRKEIEARYLQRKAAKMYTDKLETVPPLNELT VLSIIITLCCVFNCRVPRTRKEIEARYLQRKAAKMYTDKLETVPPLNELT
134 134 117 116
EVPGA---AKVEVKEEVPTVAGKVDGEKEKKKKEKESKKEGKGAKEGKEE EVPG----ALVET-SALLTVSEQVQAHAS-------------GQLPAVRE EIPGEDKKKKKK--DSVDTVAIKVEEDEK---------------NEAKKK EVPGEDKKKKKKKKDSVDTVAIKVEEDEK---------------NEAKKK
181 166 150 151
EKEEPAKKKGEKGEKGEKGAPKKEASDGGKGKKGGEKAEEKGGAKKPAKK E-DEPAN----------PGT------------------------------GEK---------------------------------------------KGEK----------------------------------------------
231 175 153 155
Zebrafish (long form) Zebrafish (short form) Mouse Human
Fig. 3. Identification of the mutation in tmie. (A) A genetic (Upper) and a physical map (Lower) depict the genomic region on chromosome 2 containing the mutation. (B) Partial chromatograms of the DNA sequence of tmie in WT (Left), homozygously mutant (Middle), and heterozygously mutant (Right) fish delineate a mutation affecting the thirteenth codon (boxed in red). (C) The gene structure of zebrafish tmie includes four coding exons, E1–E4. The short alternative transcript of tmie contains the more proximal E4a as the fourth coding exon. The asterisk denotes the location of the mutation. (D) The predicted amino acid sequence suggests a single-pass transmembrane protein after the cleavage of a signal peptide near the amino terminus (arrowhead). C, cytoplasmic surface; E, extracellular surface; M, membrane. (E) An alignment of protein sequences of Tmie from the zebrafish, mouse, and human depicts completely conserved residues in green and those identical in three instances in yellow.
induced cell death (18), we observed normal hair-cell regeneration, including robust phalloidin staining of the stereocilia and normal planar cell polarity (Fig. 2 J and K). The regenerated hair cells in the mutant nonetheless failed to incorporate 4-Di-2-ASP or FM4–64. Mapping and Identification of the Mutation. We next undertook
positional cloning to identify the mutation underlying the observed phenotypes. Using linkage analysis of over 2,300 mutant embryos for meiotic recombination, we mapped the affected locus to a genetic interval of 0.1 cM on chromosome 2 flanked by the markers ccdc13 and pthr1 (Fig. 3A). The critical region was Gleason et al.
contained entirely in the genomic insert of a fully sequenced bacterial artificial chromosome (BAC) clone, CH211–163F. After examining the critical region of this clone for potential transcripts, we found a single enticing candidate in the zebrafish ortholog of the mammalian transmembrane inner ear (Tmie) gene. This locus had escaped identification in previous zebrafish mutagenic screens for hearing defects but was recently investigated through a morpholino knockdown approach (19). Sequence analysis of the coding region of the tmie cDNA from mutant fish revealed two changes to the thirteenth triplet codon of the first coding exon, a guanosine-to-adenosine transition and the deletion of a single guanosine (Fig. 3B). The consequent frameshift results in a premature stop codon 40 bp downstream and a truncated protein of only 25 aa. The mutant allele is registered at ZFIN as tmieru1000. We performed RACE to confirm the full-length sequence and discovered a methionine codon 27 bp upstream of the previously identified translation start site (19). Incorporating this putative start site, the zebrafish tmie gene has an ORF of 693 bp comprising four coding exons that span ⬇30 kb of genomic DNA and shows strong similarity to the mouse and human genes in the structure of exon–intron boundaries (Fig. 3C). Although slightly larger, the zebrafish protein shares several features with its mammalian counterpart in topology and amino acid composition. Sequence analysis (http://bp.nuap.nagoya-u.ac.jp/sosui; 20) of the zebrafish Tmie protein predicts that it contains near its amino terminus two hydrophobic ␣-helices, the first of which likely serves as a signal peptide. If this initial hydrophobic region is cleaved, Tmie is expected to be anchored in the membrane through its second hydrophobic domain and to comprise a short extracellular or luminal segment near its amino terminus and a longer, charged cytoplasmic region at its carboxyl terminus (Fig. 3D). The products of our RACE included a second, shorter tmie transcript that reflects the use of an alternative exon 4 that lies closer to exon 3 in the genome (Fig. 3C). The resultant protein lacks the lysine-rich carboxyl terminus but contains the other conserved regions (Fig. 3E). Unless otherwise noted, the subsequent comments about the zebrafish tmie gene and corresponding protein refer to the longer forms. Confirmation of the Mutation. To confirm that the mutant phe-
notype results from the identified mutation in tmie, we microinjected RNA encoding the WT zebrafish protein into the offspring of mutant carriers and assayed rescue by 4-Di-2-ASP labeling at 4 days postfertilization (dpf). Of 40 injected embryos, three genotyped mutants were partially rescued, showing clear labeling of several neuromasts in the posterior lateral line (Fig. 4 A–C). RNA coding for the short isoform of Tmie failed to rescue mutants (results not shown). In the converse experiment, we injected WT embryos with an antisense morpholino designed to block normal splicing of the premRNA (Fig. 4 D and E). Delivery of 0.5 ng of morpholino per embryo resulted in 46% survival with consistently diminished or absent labeling of lateral-line neuromasts, although the degree of reduced labeling varied slightly between larvae. Confirming the efficacy of the morpholino by RT-PCR, we observed various aberrant splice products in both the major tmie transcript and its alternatively spliced form in injected fish (Fig. 4F). Sequence analysis of the amplicons revealed exon skipping and usage of cryptic splice sites in morphant fish. Expression Dynamics. To investigate the time course of tmie expression, we performed qRT-PCR (quantitative real-time PCR) on pools of embryos at different stages of development over the interval 24–120 hours postfertilization (hpf). Our results indicated a prominent peak of expression at 38 hpf, which represented a 390-fold increase in expression relative to the level PNAS 兩 December 15, 2009 兩 vol. 106 兩 no. 50 兩 21349
NEUROSCIENCE
A
A
B
WT
Mutant
D
C
MO
RNA
E
F
5 E1
E2
E3
MO
E4a
E4
3
L WT 1 2
3
3
0.2
0.2 E4
G
L WT 1 2 0.5 0.4 0.3
0.5 0.4 0.3
E4a
Relative expression
500 400 300 200 100 0
28 32 38 45 50 56 61 68 74 96 120 Time (hpf)
Fig. 4. Confirmation of the mutation. (A) Exposure of a WT 4-dpf larva to 4-Di-2-ASP (yellow) labels neuromasts. (B) Labeling is absent in a mutant. (C) Injection of capped RNA coding for normal Tmie partially rescues a mutant, restoring labeling in some neuromasts (arrowheads). (D) Injection of a WT embryo with a morpholino against tmie spares fluorophore incorporation in some neuromasts of the anterior lateral line but not in those of the posterior lateral line. (Scale bar, 500 m.) (E) The morpholino (MO) targets the splice junction between exon 2 and intron 2. The arrows show the locations of the primers used to amplify the splicing products in RT-PCRs. (F) The RT-PCR products amplified from one WT and three morphant fish (1-3) confirm that the injected morpholino disrupts normal splicing of tmie (Left) and of its alternative transcript (Right), producing various aberrant products. L, DNA ladder in kilobases. (G) Endogenous expression of Tmie is sparse. The relative expression of tmie at different developmental stages was measured by qRTPCR with RNA extracted from whole larvae. The results are presented as the mean and standard deviation of the ratio of each measurement to that at 24 hpf, the lowest value observed. Expression peaks at 38 hpf, declining sharply thereafter and remaining low through 120 hpf.
at 24 hpf. The second highest level of expression, which was 78-fold as great as the level at 24 hpf, was observed at 32 hpf. Expression dropped abruptly by 45 hpf and remained low at least through 120 hpf (Fig. 4G). Hair-Cell Morphology. Hair bundles gradually degenerate in spinner and circling mice, which carry mutations of the Tmie gene (21, 22). To determine whether the absence of microphonic potentials in mutant zebrafish and the inability of fluorophores to 21350 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0911632106
enter their hair cells is associated with a structural abnormality of the hair bundles, we performed scanning and transmission electron microscopy on hair cells in the posterior lateral lines of mutants. Scanning electron microscopy at 6–7 dpf detected several pathological changes in the neuromasts of mutants with respect to those of wild-type animals, including fewer and shorter kinocilia and a reduced number of stereocilia (Fig. 5 A and B). In addition, the pair of semilunar periderm cells surrounding each neuromast were often distorted. These abnormalities persisted in the neuromasts of 4-month-old homozygotes. The neuromasts of mutant larvae appeared grossly normal at the level of transmission electron microscopy (Fig. 5 C and D). Mutant hair cells displayed afferent synapses with vesicletethered ribbons and apposing postsynaptic densities as well as efferent endings. Numerous large lysosomes occurred in both hair and supporting cells. The hair bundles were significantly perturbed in all four neuromasts examined. The stereocilia of WT larvae exhibited tapering at their bases and a slight swelling at their tips; many displayed insertional plaques, and some had tip links (Fig. 5E). With no correction for shrinkage during histological preparation, the average diameter near the middle of 10 stereociliary shafts was 142 ⫾ 20 nm (mean ⫾ standard deviation). In mutant larvae, by contrast, the stereocilia lacked both insertional plaques and tip links (Fig. 5F). Ten of these stereocilia averaged only 87 ⫾ 22 nm in diameter. Discussion The mutant phenotype in zebrafish accords with the startle and balance deficits in spinner and circling mice and with the profound deafness in humans affected at the DFNB6 locus (23). Despite its obvious importance in auditory and vestibular function, the role of Tmie in mechanoelectrical transduction remains a mystery. Our results with mutant zebrafish indicate that a problem exists at the level of hair cells, in which severe impairments in activity were detected by electrical recordings and the absence of fluorophore uptake. In contrast to a recent study of Tmie in zebrafish using morpholino knockdown (19), we did not observe in mutant zebrafish smaller eyes or inner ears, abnormalities in semicircular-canal development, disturbances in otolith formation, delayed lateral-line patterning, or an obvious reduction in the number of phalloidin-positive hair cells. These phenotypic consequences might have represented nonspecific effects associated with morpholino injections, which we also observed at high concentrations. Tmie bears little resemblance to any proteins of known function. A yeast-two hybrid analysis identified the protein inhibitor of activated STAT3 (PIAS3) as a potential binding partner, but our attempts to verify the interaction through pull-down assays and coimmunoprecipitation were unsuccessful. Although hair-cell activity in the mutants is compromised, it is unclear whether tmie is expressed in hair cells or in the adjacent supporting cells. Our efforts to localize the protein through in situ hybridization and immunohistochemistry were unsuccessful, probably owing to the low level of expression. This low expression accords with our qRT-PCR results, which revealed that transcript levels in the whole animal peak briefly early in development but then plummet. Although the brief surge in expression might shed light on the function of Tmie, the time at which expression peaks coincides with no known developmental milestone and is therefore uninformative at present. We obtained signals from immunolabeling of zebrafish only when Tmie was overexpressed as an eGFP fusion protein. Consistent with predictions of a transmembrane domain in Tmie, labeling was associated with the plasma membrane. It is likely that the putatively cytoplasmic region of Tmie is essential to its function. Human mutations that disrupt arginine residues there lead to profound hearing loss (23, 24). This highly charged Gleason et al.
A
B
E E
200 nm
F 5 µm
D
C
Fig. 5. Ultrastructural features of mutant and normal hair cells. (A) A scanning electron micrograph of a midtail neuromast from the posterior lateral line of a mutant at 6 dpf shows a dozen long kinocilia emerging from short clusters of stereocilia. (B) A neuromast from a similar position in an age-matched, homozygously mutant fish displays shrunken kinocilia and a diminished number of hair bundles. The semilunar periderm cells surrounding the neuromast are abnormally smooth. (C) A low-power transmission electron micrograph of a neuromast in a 6-dpf control larva shows parts of seven hair cells, some contacting large nerve terminals. (D) A mutant neuromast is characterized by several lysosomes (arrowheads) both within them and in supporting cells. A representative lysosome is enlarged fivefold in the inset. (E) A high-power micrograph from a control animal demonstrates a pair of sterocilia connected by a tip link (arrowhead) that terminates at its upper end in a prominent insertional plaque. (F) In a mutant larva of the same age, the stereocilia are narrow and display little basal tapering. Neither tip links nor insertional plaques are evident, and the shorter stereocilia lack dense material beneath their tips. The magnifications are identical for A and B, for C and D, and for E and F.
region could serve, for example, as a site of interaction with other proteins that have a direct role in mechanotransduction. The mutant’s deficiencies in microphonic responsiveness and fluorophore uptake point to a problem with the mechanoelectrical-transduction apparatus. Such a defect seems still more likely in view of the electron-microscopic study, which revealed that the stereocilia of mutants appear immature and lack both tip links and the insertional plaques at their upper ends. The deficiency in transduction might result from the presence of immature hair cells, perhaps because older cells degenerate. The defect might alternatively reflect the mislocalization, misassembly, or absence of the mechanotransduction machinery. In its extensive content of basic residues, Tmie resembles CALF-1, a protein of Caenorhabditis thought to constitute a subunit of or the chaperone for a voltage-gated Ca2⫹ channel (25). It is possible that Tmie serves a similar function for the hair cell’s mechanosensitive channel or for another essential component of the transduction apparatus. Materials and Methods Zebrafish Husbandry and Discovery of the Mutants. Adult zebrafish were maintained and bred according to standard protocols (26). Fluorophore Labeling of Hair Cells and Electrical Recordings. To label lateralline neuromasts, we immersed larvae for 1–2 min in a solution of 200 M 4-Di-2-ASP or 3 M FM4 – 64 (Invitrogen) in breeding water at room temperature, then in several changes of fluorophore-free water. Labeled fish were imaged under epifluorescence illumination with a 5⫻ or 10⫻ objective lens and a charge-coupled-device camera (DP-71, Olympus).
Gleason et al.
To label hair cells of the inner ear, we used microelectrodes to pressureinject fluorophores in 120 mM KCl into the otic cavities of larvae immobilized in 1.2% agar. Microphonic potentials were measured from the otic vesicles of zebrafish larvae at 5– 6 dpf as described (9) except that the recording pipette contained 120 mM KCl and the stimulator provided 200-Hz sinusoidal displacements of ⫾2 m. Hair-Cell Regeneration and Phalloidin Staining. Live 5-dpf WT and mutant Et(krt4:GFP)sqet4 larvae were immersed in 5 M CuSO4 in E3 medium (18) for 1 h at 28.5 °C and then rinsed twice briefly in E3 medium. After confirming essentially complete hair-cell death in the lateral line by fluorescence microscopy, we allowed treated animals to recover in E3 medium at 28.5 °C. Hair-cell regeneration was examined 50 h later by confocal microscopy, after which the animals were fixed for staining. Filamentous actin was stained by phalloidin as described (6). Positional Cloning and Identification of the Mutation. A map cross was performed by mating adult mutant carriers to WT animals obtained from a pet store. After the progeny had been raised and inbred, their offspring were screened for the mutant phenotype by behavior or fluorophore labeling. Bulked segregant analysis (27) with a set of 214 microsatellite markers was performed on the genomic DNA extracted from pools of 20 larvae to identify chromosomal linkage. For high-resolution mapping, more than 2,300 mutants were analyzed to assess the frequency of meiotic recombination with respect to published microsatellite markers and single-nucleotide polymorphisms that had been identified by genomic DNA sequencing. To generate a physical map, we used markers flanking the critical interval to search the database of sequenced BAC clones (http://www.sanger.ac.uk/Projects/D㛭rerio). The sequence of clone CH211–163F was examined for candidate transcripts with BLAST (http://blast.ncbi.nlm.nih.gov/Blast.cgi). The cDNA and genomic sePNAS 兩 December 15, 2009 兩 vol. 106 兩 no. 50 兩 21351
NEUROSCIENCE
5 µm
quences of tmie obtained from WT and mutant animals were then compared to reveal the mutation. RACE. Total RNA extracted from 5-dpf larvae was used for synthesis of adaptortagged cDNA (SMART, Clontech). RACE PCR was performed according to the manufacturer’s protocol with gene-specific primers targeting exon 1 (for 5⬘ RACE) or exon 3 (for 3⬘ RACE). Microinjection. Pressure-injection (PLI-100, Warner Instruments) was performed on one-cell-stage embryos using solutions in nuclease-free water. For rescue experiments, embryos were injected with 100 pg of capped RNA synthesized (mMESSAGE mMACHINE, Ambion) to accord with the coding region of zebrafish tmie. For phenocopy experiments, embryos received 500 pg of a morpholino (5⬘-GCAATTACACTTGCAGAAACTTACT-3⬘) designed to block the splice junction between exon 2 and intron 2 in the premRNA of both the long and short transcripts. We assessed the extent of 4-Di-2-ASP labeling in injected larvae at 4 dpf.
qRT-PCR was performed (ABI Prism 7700, Applied Biosystems) with a melting temperature of 59 °C. Three pools of embryos were processed for each experimental time point, with each reaction run in triplicate. To measure relative quantification of transcripts, we used the standardcurve method (Applied Biosystems Prism SDS 7700 User Bulletin 2). For each primer set, we used serial dilutions of cDNA to generate a calibration curve that then allowed extrapolation of relative expression levels of tmie at different times. These values were first normalized to amplification of rpl13 alpha (29), an endogenous control. We finally expressed values as multiples of those obtained from 24-hpf samples, which showed the lowest level of expression. The GenBank accession number for the long splice isoform of tmie is GQ389648 and that for the short isoform is GQ389649.
qRT-PCR. Total RNA was extracted from pools of 10 embryos (TRIzol, Invitrogen) for synthesis of random-hexamer-primed cDNA (Applied Biosystems).
ACKNOWLEDGMENTS. The authors thank H. Baier for Tg(pou4f3:gap43mGFP) zebrafish, D. Gilmour for Tg(cldnB:lynGFP) zebrafish, V. Korzh for Et(krt4:GFP)sqet4 zebrafish, and J. Y. Kuwada for Tg(Hsp70:eGFP) zebrafish. We also thank A. Afolalu for excellent zebrafish husbandry and B. Fabella for computer programming and technical assistance with microphonic recordings. Members of our research group provided valuable comments on the manuscript. This investigation was supported by grants DC00241 and GM07739 from the National Institutes of Health and by grant SAF2006 – 04684 from the Ministerio de Ciencia e Innovacio´n of Spain, of which H.L.-S. is a Ramo´n y Cajal Investigator. A.N. is the recipient of a Ruth L. Kirschstein National Research Service Award Predoctoral Fellowship. M.R.G. is an associate and A.J.H. is an investigator of the Howard Hughes Medical Institute.
1. Hudspeth AJ (1989) How the ear’s works work. Nature 341:397– 404. 2. Petit C, Levilliers J, Hardelin JP (2001) Molecular genetics of hearing loss. Annu Rev Genet 35:589 – 646. 3. Dooley K, Zon LI (2000) Zebrafish: A model system for the study of human disease. Curr Opin Genet Dev 10:252–256. 4. Haddon C, Lewis J (1996) Early ear development in the embryo of the zebrafish, danio rerio. J Comp Neurol 365:113–128. 5. Blaxter JHS, Fuiman LA (1989) in The Mechanosensory Lateral Line: Neurobiology and Evolution, eds Coombs S, Go¨rner P, Mu¨nz H (Springer, New York), pp 481– 499. 6. Asai Y, et al. (2006) Mutation of the atrophin2 gene in the zebrafish disrupts signaling by fibroblast growth factor during development of the inner ear. Proc Natl Acad Sci USA 103:9069 –9074. 7. Haffter P, Nusslein-Volhard C (1996) Large scale genetics in a small vertebrate, the zebrafish. Int J Dev Biol 40:221–227. 8. Kappler JA, Starr CJ, Chan DK, Kollmar R, Hudspeth AJ (2004) A nonsense mutation in the gene encoding a zebrafish myosin vi isoform causes defects in hair-cell mechanotransduction. Proc Natl Acad Sci USA 101:13056 –13061. 9. Starr CJ, Kappler JA, Chan DK, Kollmar R, Hudspeth AJ (2004) Mutation of the zebrafish choroideremia gene encoding rab escort protein 1 devastates hair cells. Proc Natl Acad Sci USA 101:2572–2577. 10. Blechinger SR, et al. (2002) The heat-inducible zebrafish hsp70 gene is expressed during normal lens development under non-stress conditions. Mech Dev 112:213–215. 11. Foreman MB, Eaton RC (1993) The direction change concept for reticulospinal control of goldfish escape. J Neurosci 13:4101– 4113. 12. Gale JE, Marcotti W, Kennedy HJ, Kros CJ, Richardson GP (2001) FM1– 43 dye behaves as a permeant blocker of the hair-cell mechanotransducer channel. J Neurosci 21:7013– 7025. 13. Meyers JR, et al. (2003) Lighting up the senses: FM1– 43 loading of sensory cells through nonselective ion channels. J Neurosci 23:4054 – 4065. 14. Xiao T, Roeser T, Staub W, Baier H (2005) A gfp-based genetic screen reveals mutations that disrupt the architecture of the zebrafish retinotectal projection. Development 132:2955–2967. 15. Nicolson T, et al. (1998) Genetic analysis of vertebrate sensory hair cell mechanosensation: The zebrafish circler mutants. Neuron 20:271–283.
16. Haas P, Gilmour D (2006) Chemokine signaling mediates self-organizing tissue migration in the zebrafish lateral line. Dev Cell 10:673– 680. 17. Lo´pez-Schier H, Hudspeth A (2006) A two-step mechanism underlies the planar polarization of regenerating sensory hair cells. Proc Natl Acad Sci USA 103:18615–18620. 18. Hernandez PP, Olivari FA, Sarrazin AF, Sandoval PC, Allende ML (2007) Regeneration in zebrafish lateral line neuromasts: Expression of the neural progenitor cell marker sox2 and proliferation-dependent and-independent mechanisms of hair cell renewal. Dev Neurobiol 67:637– 654. 19. Shen YC, et al. (2008) The transmembrane inner ear (tmie) gene contributes to vestibular and lateral line development and function in the zebrafish (Danio rerio). Dev Dyn 237:941–952. 20. Hirokawa T, Boon-Chieng S, Mitaku S (1998) Sosui: Classification and secondary structure prediction system for membrane proteins. Bioinformatics 14:378 –379. 21. Mitchem KL, et al. (2002) Mutation of the novel gene tmie results in sensory cell defects in the inner ear of spinner, a mouse model of human hearing loss dfnb6. Hum Mol Genet 11:1887–1898. 22. Chung WH, et al. (2007) Cochlear pathology of the circling mouse: A new mouse model of dfnb6. Acta Otolaryngol 127:244 –251. 23. Naz S, et al. (2002) Mutations in a novel gene, tmie, are associated with hearing loss linked to the dfnb6 locus Am J Hum Genet 71:632– 636. 24. Santos RL, et al. (2006) Novel sequence variants in the tmie gene in families with autosomal recessive nonsyndromic hearing impairment. J Mol Med 84:226 –231. 25. Saheki Y, Bargmann CI (2009) Presynaptic CaV2 calcium channel traffic requires CALF-1 and the ␣2␦ subunit UNC-36. Nat Neurosci 12:1257–1265. 26. Westerfield M (2000) The Zebrafish Book. A Guide for the Laboratory Use of Zebrafish (Danio rerio), Ed 4 (University of Oregon Press, Eugene). 27. Postlethwait JH, Talbot WS (1997) Zebrafish genomics: From mutants to genes. Trends Genet 13:183–190. 28. Nagiel A, Andor-Ardo D, Hudspeth AJ (2008) Specificity of afferent synapses onto plane-polarized hair cells in the posterior lateral line of the zebrafish. J Neurosci 28:8442– 8453. 29. Tang R, Dodd A, Lai D, McNabb WC, Love DR (2007) Validation of zebrafish (Danio rerio) reference genes for quantitative real-time rt-pcr normalization. Acta Biochim Biophys Sin (Shanghai) 39:384 –390.
Electron Microscopy. For transmission electron microscopy, 6-dpf and 9-dpf mutant and WT larvae were prepared as described (28). For scanning electron microscopy, 6-dpf and young adult mutant and WT fish were processed as described (9) except that the specimens were coated with gold palladium before examination at an accelerating voltage of 5 kV in a scanning electron microscope (1550, LEO Electron Microscopy) with a 30-m aperture.
21352 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0911632106
Gleason et al.
