Putative pore-forming subunits of the mechano-electrical transduction channel, Tmc1/2b, require Tmie to localize to the site of mechanotransduction in zebrafish sensory hair cells

Mutations in transmembrane inner ear (TMIE) cause deafness in humans; previous studies suggest involvement in the mechano-electrical transduction (MET) complex in sensory hair cells, but TMIE’s precise role is unclear. In tmie zebrafish mutants, we observed that GFP-tagged Tmc1 and Tmc2b, which are putative subunits of the MET channel, fail to target to the hair bundle. In contrast, overexpression of Tmie strongly enhances the targeting of Tmc2b-GFP to stereocilia. To identify the motifs of Tmie underlying the regulation of the Tmcs, we systematically deleted or replaced peptide segments. We then assessed localization and functional rescue of each mutated/chimeric form of Tmie in tmie mutants. We determined that the first putative helix was dispensable and identified a novel critical region of Tmie, the extracellular region and transmembrane domain, which mediates both mechanosensitivity and Tmc2b-GFP expression in bundles. Collectively, our results suggest that Tmie’s role in sensory hair cells is to target and stabilize Tmc subunits to the site of MET. Author summary Hair cells mediate hearing and balance through the activity of a pore-forming channel in the cell membrane. The transmembrane inner ear (TMIE) protein is an essential component of the protein complex that gates this so-called mechanotransduction channel. While it is known that loss of TMIE results in deafness, the function of TMIE within the complex is unclear. Using zebrafish as a deafness model, Pacentine and Nicolson demonstrate that Tmie is required for the localization of other essential complex members, the transmembrane channel-like (Tmc) proteins, Tmc1/2b. They then evaluate twelve unique versions of Tmie, each containing mutations to different domains of Tmie. This analysis reveals that some mutations in Tmie cause dysfunctional gating of the channel as demonstrated through reduced hair cell activity, and that these same dysfunctional versions also display reduced Tmc expression at the normal site of the channel. These findings link hair cell activity with the levels of Tmc in the bundle, reinforcing the currently-debated notion that the Tmcs are the pore-forming subunits of the mechanotransduction channel. The authors conclude that Tmie, through distinct regions, is involved in both trafficking and stabilizing the Tmcs at the site of mechanotransduction.


46
The auditory and vestibular systems detect mechanical stimuli such as sound, gravity, and 47 acceleration. These two systems share a sensory cell type called hair cells. The somas of hair 48 cells are embedded in the epithelium and extend villi-like processes from their apex into the 49 surrounding fluid. The shorter of these, the stereocilia, are arranged in a staircase-like pattern 50 adjacent to a single primary cilium known as a kinocilium. Neighboring cilia are connected by 51 protein linkages. Deflection of the kinocilium along the excitatory axis tugs the interconnected 52 stereocilia, which move as a single unit called the hair bundle [1]. When tension is placed on the 53 upper-most linkages known as tip links, the force is thought to open mechanosensitive channels 54 at the distal end of the shorter stereocilia [2,3]. These channels pass current, depolarizing the 55 cell and permitting electrical output to the brain via the eighth cranial nerve. The conversion of a 56 mechanical stimulus into an electrical signal is known as mechano-electrical transduction (MET) 57 [4]. The proteins located at the site of MET and involved in gating the MET channel are 58 8 (Fig 1E and 1F). Tmie-GFP also restores microphonic potentials to wild-type levels ( Fig 1H,  150 orange trace). In a stable line with a single transgene insertion, we observed that Tmie-GFP 151 expression varies among hair cells, even within the same patch of neuroepithelium (lateral crista, 152 S1A Fig). Immature hair cells, which can be identified by their shorter stereocilia and kinocilia 153 (S1A Fig, bracket and  Having confirmed that our exogenously expressed Tmie-GFP is functional, we used this 166 transgene to probe Tmie's role in the MET complex. First, we characterized Tmie's interactions 167 with other MET proteins in vivo by expressing transgenic Tmie-GFP in mutant pcdh15a, lhfpl5a, 168 and tomt larvae (Fig 2). Because a triple knock-out of zebrafish tmc has not been reported, we 169 used tomt mutants as a proxy for tmc-deficient fish based on recent studies of defective bundle 170 localization of the Tmcs in tomt-deficient fish and mice [11,35]. As in wild type bundles ( Fig  171   2A), Tmie-GFP is detectable in the stereocilia in each of these MET mutants (Fig 2B and 2D), 9 even if hair bundles are splayed (Fig 2B and 2C, arrowheads). This result suggests that Tmie 173 does not depend on interactions with other MET components for entry into the hair bundle. 174 175 Tmc1-GFP and Tmc2b-GFP fail to localize to stereocilia without Tmie 176 To determine if the loss of Tmie affects the other components of the 177 mechanotransduction complex, we expressed GFP-tagged mechanotransduction proteins 178 (Pcdh15aCD3, Lhfpl5a, Tmc1, and Tmc2b) in tmie ru1000 mutants. Both Pcdh15aCD3-GFP (Fig  179   3A) and GFP-Lhfpl5a ( Fig 3B) showed GFP fluorescence in hair bundles with a punctate 180 distribution, similar to the pattern seen in wild type bundles. This result is consistent with the 181 intact morphology of tmie ru1000 hair bundles. However, when we imaged Tmc1-GFP ( Fig 3C) and 182 Tmc2b-GFP (Fig 3E), GFP fluorescence was severely reduced in the hair bundles of tmie ru1000 183 mutants. In mature tmie ru1000 hair cells, we often saw a signal within the apical soma near the 184 cuticular plate, indicative of a trafficking defect (Fig 3E, arrows; position of cuticular plate 185 denoted in Fig 1G). We quantified Tmc expression in the hair bundle region and observed a 186 striking and consistent reduction in tmie mutants (Fig 3D and 3F p2A-NLS(mCherry) driven by the myo6b promoter. The p2A linker is a self-cleaving peptide, 195 which leads to translation of equimolar amounts of Tmie and NLS(mCherry). Hence, mCherry 196 expression in the nucleus denotes Tmie expression in the cell (Fig 3G, lower panels). We 197 generated a stable tmie ru1000 fish line carrying the tmie-p2A-NLS(mCherry) transgene and then 198 crossed it to the Tg(myo6b:tmc2b-GFP); tmie ru1000 line. We observed that overexpression of 199 Tmie led to a 2.5-fold increase in expression of Tmc2b-GFP in the bundles of hair cells when 200 compared to wild type siblings that carried only the tmc2b-GFP transgene (Fig 3G and 3H). 201 Combined with the finding that Tmc expression is lost in hair bundles lacking Tmie, our data 202 suggest that Tmie positively regulates Tmc localization to the hair bundle. 203 204

Transgenes can effectively determine protein functionality 205
To gain a better understanding of Tmie's role in regulating the Tmcs, we characterized a 206 new allele of tmie, t26171, which was isolated in a forward genetics screen for balance and 207 hearing defects in zebrafish larvae. Sequencing revealed that tmie t26171 fish carry an A→G 208 mutation in the splice acceptor of the final exon of tmie, which leads to use of a nearby cryptic 209 splice acceptor (S2A Fig, DNA, cDNA). Use of the cryptic acceptor causes a frameshift that 210 terminates the protein at amino acid 139 (A140X), thus removing a significant portion of the C-211 terminal tail (S2A Fig, Protein). Homozygous mutant larvae exhibit severe auditory and 212 vestibular deficits, being insensitive to acoustic stimuli and unable to maintain balance (S2A Fig,  213 Balance). FM 4-64 labeling of tmie t26171 mutant hair cells suggests that the effect of the mutation 214 is similar to the ru1000 mutation (S2B and S2D Figs). This finding implicates the C-terminal 215 tail, a previously uncharacterized region, in Tmie's role in MET. However, when we 216 overexpressed a near-mimic of the predicted protein product of tmie t26171 (1-138-GFP) using the 217 myo6b promoter, we observed full rescue of FM labeling defects in tmie ru1000 (S2C and S2D 218 Figs), as well as behavioral rescue of balance and acoustic sensitivity (n=19). These results 219 revealed that when expressed at higher levels, loss of residues 139-231 does not have a 220 significant impact on Tmie's ability to function. 221 This paradoxical finding highlighted an important advantage of the use of transgenes over 222 traditional mutants. There are myriad reasons why a genomic mutation may lead to dysfunction, 223 including reduced transcription or translation, protein misfolding and degradation, or 224 mistrafficking. Exogenous expression may overcome these deficiencies by producing proteins at 225 higher levels. Moreover, the use of transgenes enabled us to carry out a more comprehensive 226 structure/function study of Tmie. To test a collection of deletions and chimeras of Tmie, we 227 therefore used the myo6b promoter to drive exogenous expression of the constructs in hair cells 228 of the tmie ru1000 mutant. 229 We systematically deleted or replaced regions of tmie to generate 12 unique tmie 230 constructs ( Fig 4A). Earlier studies in zebrafish and mice proposed that Tmie undergoes 231 cleavage, resulting in a single-pass mature protein [21,36]. To test this hypothesis, we generated 232 the SP44-231 construct of Tmie, which replaced the N-terminus with a known signal peptide 233 (SP) from a zebrafish Glutamate receptor protein (Gria2a). The purpose of the unrelated signal 234 peptide was to preserve the predicted membrane topology of Tmie. We also made a similar 235 construct that begins at amino acid 63, where the sequence of Tmie becomes highly conserved 236 (SP63-231). Three of the constructs contained internal deletions (63-73; 97-113; 114-138).

237
In three more constructs, we replaced part of or the entire second transmembrane helix (2TM) 238 with a dissimilar helix from the CD8 glycoprotein (CD8; CD8-2TM; 2TM-CD8). We included 239 our mimic of the zebrafish tmie t26171 mutant, which truncates the cytoplasmic C-terminus (1-240 138). To further truncate the C-terminus, we made a construct that mimics the mouse sr J mutant 12 (1-113). In mice, this truncation recapitulates the full-deletion phenotype [20]. Finally, we 242 included an alternate isoform of Tmie that uses a different final exon, changing the C-terminal 243 sequence (Tmie-short). This isoform is found only in zebrafish [21] and its function has not been 244 values (bundle/bundle + soma) and expressed this as a ratio ( Fig 4D). Values closer to 1 are 256 bundle enriched, while values closer to 0 are soma-enriched. We excluded the CD8-GFP 257 construct from further analyses because it was detected only in immature bundles (Fig 4B, CD8). Bath applied FM dye demonstrates the presence of permeable MET channels, but does 293 not reveal any changes in mechanically evoked responses in hair cells. Therefore, we also 294 recorded microphonics of mutant larvae expressing individual transgenes. For our recordings, we 295 inserted a recording pipette into the inner ear cavity of 3 dpf larvae and pressed a glass probe 296 against the head (Fig 6A). Using a piezo actuator to drive the probe, we delivered a step stimulus 297 at increasing driver voltages while recording traces in current clamp ( Fig 6B). For each 298 transgenic tmie line, we measured the amplitude of the response at the onset of stimulus ( Fig 6C-299 I). We limited our analysis to the lines expressing constructs that failed to fully rescue FM 300 labeling (Fig 6E-I). As positive controls, we used the full-length tmie line (Fig 65 C) and also 301 included SP44-231 (Fig 6D), encoding the cleavage product mimic. Both controls fully rescued 302 the responses in tmie ru1000 larvae. Consistent with a reduction in labeling with FM dye, we found 303 that the microphonic responses were strongly or severely reduced in larvae expressing the SP63-304 231, 63-73, CD8-2TM, 2TM-CD8 and 97-113 constructs in the tmie ru1000 background (Fig 6E-305 I). We also saw the same dominant negative effect in wild type larvae expressing transgenic 306 SP63-231 or 63-73.

Tmc2b-GFP 310
After identifying functional regions of Tmie, we asked whether these regions are 311 involved in regulating Tmc localization. Therefore we quantified hair bundle expression of 312 transgenic Tmc2b-GFP in hair cells of tmie ru1000 mutant larvae stably co-expressing individual 313 transgenic tmie constructs (Fig 7B-H). As in Fig 3 G, we tagged our tmie constructs with p2A-314 NLS(mCherry) so that Tmc2b-GFP expression in the hair bundles could be imaged separately. 315 We examined SP44-231 and the five tmie constructs that yielded impaired mechanosensitivity. 316 Three constructs showed full rescue of Tmc2b-GFP levels in the bundle. The SP44-231 317 cleavage mimic produced highly variable levels, some in the wild type range, others increasing 318 Tmc2b-GFP expression above wild type (Figs 7B and 7H), as seen in overexpression of full-319 length Tmie (Figs 3H and 7A, right panel). We suspect that the exogenous Gria2a signal peptide 320 leads to variable processing of Tmie and thus contributes to this variability in Tmc2b-GFP 321 fluorescence. tmie ru1000 larvae expressing the SP63-231 construct gave rise to values of Tmc2b-322 GFP fluorescence within the wild type range (Fig 7C and 7H). When we recorded microphonics 323 in these larvae, we found that co-overexpression of Tmc2b-GFP and SP63-231 resulted in better 324 functional rescue of tmie ru1000 (S3A Fig) than when SP63-231 was expressed alone (Fig 6E). We 325 also determined that the microphonic potentials correlated with the levels of Tmc2b-GFP in the 326 bundles (S3B Fig). Likewise, the 2TM-CD8 construct also generated values of Tmc2b-GFP 327 fluorescence in the wild type range (Fig 7F and 7H). These larvae rescued microphonic 328 potentials to wild type levels (S3C Fig), unlike when 2TM-CD8 was expressed alone (Fig 6E). Of the three constructs with little to no functional rescue, CD8-2TM (Fig 7E and 7H) and 332 97-113 (Fig 7G and 7H) had severely reduced levels of Tmc2b-GFP in hair bundles. In 333 tmie ru1000 expressing 63-73, there was severely reduced but still faintly detectable Tmc2b-GFP 334 signal though, as with the functional rescue, this difference was not statistically significant ( Fig  335   7D and 7H). The bulk of this signal was observed in immature bundles (Fig 7D, arrows,  we examined the localization of Tmie-GFP in mutants of essential MET genes: the tip-link 371 protein pcdh15a; the accessory protein lhfpl5a; and in tomt mutants (Fig 3) Our results reveal that Tmie has distinct regions associated with self-localization and 383 function (Fig 8). Three constructs showed impaired targeting of Tmie to the bundle, namely 384 SP63-231, CD8-2TM, and the 1-138 construct, the last of which truncates the C-terminus. 385 Further manipulation to the C-terminus, either by removing more amino acids (1-113) or by 386 using an alternative final exon (Tmie-short), results in targeting of the protein to the plasma 387 membrane instead of the bundle. However, removing just a smaller internal segment has no 388 effect on bundle localization (Fig 4A and 4C, 114-138). We suspect that the abundance of 389 charged residues in the C-terminus of Tmie (S2A Fig, Protein), as well as the regions altered in 390 SP63-231 and CD8-2TM, contribute to recognition by bundle trafficking machinery. 391 Mislocalization, however, did not necessarily correlate with functional rescue. Despite partial 392 mislocalization to the plasma membrane, the 1-138 construct showed full functional rescue. 393 Conversely, despite normal localization to the bundle, 97-113 did not rescue function at all. 394 These results demonstrate that Tmie's functional role is separate from its ability to target to the 395 bundle. 396 397 Tmie promotes the levels of Tmc1/2 in the hair bundle 398 The regulatory role of Tmie with respect to the Tmcs is strongly supported by the 399 strikingly different effects of loss of Tmie versus overexpression of Tmie. When Tmie is absent, 400 19 so are the Tmcs; when Tmie is overexpressed, the level of Tmc2b in the bundle is boosted as 401 well (Fig 3C-H) Tmc2b-GFP bundle expression to wild type levels or higher. To our knowledge, these results are 425 the first in vivo evidence that Tmie can function without the putative first transmembrane 426 domain. Our study supports the notion that Tmie undergoes cleavage, resulting in a single-pass 427 membrane protein that functions in the MET complex (Fig 8B). When co-expressed with Tmc2b-GFP, our Tmie constructs reveal a strong link between 462 function and Tmc bundle expression (Figs 7 and S3). In larvae expressing CD8-2TM and 97-463 113, both of which display little or no functional rescue, there is no detectable Tmc2b-GFP in the 464 hair bundle (Fig 7E, 7G and 7H). In addition to defects in targeting Tmcs to the hair bundle, our 465 data also suggest a role for Tmie in maintaining the levels of Tmc2b in stereocilia. The most 466 dramatic effect on maintenance of Tmc signal in the bundle was seen in tmie ru1000 larvae 467 expressing the 63-73 construct. In these larvae, Tmc2b-GFP successfully traffics to the bundle 468 in immature hair cells (Fig 7D, arrows) (Fig 8B). 472 Surprisingly, SP63-231 and 2TM-CD8 rescue Tmc2b-GFP to wild type levels (Fig 7C,  473 7F and 7H), even though functional rescue of tmie ru1000 by GFP-tagged versions was reduced in 474 both FM labeling experiments (Fig 5) and microphonic recordings of the inner ear (Fig 6). This 475 result hints at an additional role for Tmie in MET that is independent of Tmc trafficking.  Table 1. 505 same as pME-CD8-2TM pME-∆97-113 AGACTTGCTGCGAAAAATTATGCCAAC GAAGATGCAGCAGAGCGTAATTATTATTGC pME-∆114-138 GCGGCAAAGGTTGAGGTGAAG TTGCGCGTGCCGAGC pME-1-138 same as pME-Tmie GGGGACCACTTTGTACAAGAAAGCTGGGTC GCCGGGCACCTCAG pME-1-113 same as pME-Tmie GGGGACCACTTTGTACAAGAAAGCTGGGTC TTGCGCGTGCCGAG pME-Tmie-short same as pME-Tmie GGGGACCACTTTGTACAAGAAAGCTGGGTC AGTGCCAGGATTGGCTG The 5' entry vector contained the promoter for the myosin 6b gene, which drives 521 expression only in hair cells. All tmie transgenic constructs were subcloned into the middle entry 522 vector using PCR or bridging PCR and confirmed by sequencing. The primers for each vector 523 are listed in Table 1. For GFP-tagging, we used a 3' entry vector with a flexible linker 524 Experiments were conducted as previously described [49]. Wild type and mutant larvae 575 were sorted by FM 1-43 labeling. Briefly, 6 dpf larvae were placed in six central wells of a 96-576 well microplate mounted on an audio speaker. Pure tones were played every 15 s for 3 min 577 (twelve 100 ms stimuli at 1 kHz, sound pressure level 157 dB, denoted by asterisks in Fig 1B). 578

Primers for RT-PCR
Responses were recorded in the dark inside a Zebrabox monitoring system (ViewPoint Life 579 Sciences). Peaks represent pixel changes from larval movement. A response was considered 580 positive if it occurred within two seconds after the stimulus and surpassed threshold to be 581 considered evoked, not spontaneous ( Fig 1B, green indicates movement detected, magenta 582 indicates threshold surpassed). For each larva, we used the best response rate out of three trials. 583 Response was quantified by dividing the number of positive responses by total stimuli (12) and 584 converting to a percent. If the larvae moved within two seconds before a stimulus, that stimulus 585 was dropped from the trial data set (i.e. the number of total stimuli would become 11). Each data 586 point on the graph in Fig 1C is  for 25-30 seconds, larvae were washed 3x in E3. Laser power was adjusted for each experiment 595 to avoid saturation of pixels but was consistent within a clutch. FM levels were quantified in 596 ImageJ [50] as described previously [9]. In brief, maximum projections of each neuromast were 597 generated using seven optical sections, beginning at the cuticular plate and moving down through 598 the soma (magenta bracket, Fig 1G). We then measured the integrated density of the channel 599 with an emission peak at 640 nm for FM 4-64, and at 488 nm for FM 1-43. This integrated 600 pipettes to a resistance of 3-6 MΩ. We pulled a second pipette to a long shank and fire polished 617 to a closed bulb, and then attached this rod to a piezo actuator (shielded with tin foil). The rod 618 was then pressed to the front of the head behind the lower eye, level with the otoliths in the ear 30 of interest, to hold the head in place while the recording pipette was advanced until it pierced the 620 inner ear cover. Although it has been demonstrated that size of response is unchanged by entry 621 point [51], we maintained a consistent entry point dorsal to the anterior crista and lateral to the 622 posterior crista (see Fig 6A). After the recording pipette was situated, the piezo pipette was then 623 moved back to a position in light contact with the head. We drove the piezo with a High Power 624 Amplifier (piezosystem jena, System ENT/ENV, serial # E18605), and recorded responses in 625 current clamp mode with a patch-clamp amplifier (HEKA, EPC 10 usb double, serial # 550089). hand-drawn area (Fig 7A, right panel, black outline). In the ROI, we quantified the integrated 644 density of the channel with an emission peak at 480 nm. This was repeated in the region above 645 the bundles containing only inner ear fluid and the kinocilia in order to subtract background 646 fluorescence. Each middle crista generated one data point on the graphs in Figs 3 and 7. In some 647 cases, we saw single cells that appeared to have a GFP-fill, probably due to clipping of the GFP 648 tag. We excluded these cells from analyses, since they falsely increased the signal. Due to the 3D 649 nature of the mound-shaped cristae, it was difficult to completely exclude the apical soma region, 650 leading the signals of tmie ru1000 to average above zero. We used the Kruskal-Wallis test for the 651 SP44-231, SP63-231, and 2TM-CD8 constructs; all others are one-way ANOVA. We sorted 30 wild type and 30 t26171 larvae by behavior (tap sensitivity and balance 655 defect at 5 dpf) and extracted RNA using the RNeasy mini kit (Qiagen). Larvae were 656 homogenized using a 1ml syringe. To generate the cDNA for the short isoform of Tmie (Tmie-657 short) and the t26171 allele, we performed RT-PCR on these RNA samples using the RNA to 658 cDNA EcoDry Premix (Clontech, Cat # 639549). Primers are listed in Table 1