Mechanical overstimulation causes acute injury and synapse loss followed by fast recovery in lateral-line neuromasts of larval zebrafish

Excess noise damages sensory hair cells, resulting in loss of synaptic connections with auditory nerves and, in some cases, hair-cell death. The cellular mechanisms underlying mechanically induced hair-cell damage and subsequent repair are not completely understood. Hair cells in neuromasts of larval zebrafish are structurally and functionally comparable to mammalian hair cells but undergo robust regeneration following ototoxic damage. We therefore developed a model for mechanically induced hair-cell damage in this highly tractable system. Free swimming larvae exposed to strong water wave stimulus for 2 hr displayed mechanical injury to neuromasts, including afferent neurite retraction, damaged hair bundles, and reduced mechanotransduction. Synapse loss was observed in apparently intact exposed neuromasts, and this loss was exacerbated by inhibiting glutamate uptake. Mechanical damage also elicited an inflammatory response and macrophage recruitment. Remarkably, neuromast hair-cell morphology and mechanotransduction recovered within hours following exposure, suggesting severely damaged neuromasts undergo repair. Our results indicate functional changes and synapse loss in mechanically damaged lateral-line neuromasts that share key features of damage observed in noise-exposed mammalian ear. Yet, unlike the mammalian ear, mechanical damage to neuromasts is rapidly reversible.


: Intense water wave produced by shaker apparatus stimulates lateral-line hair cells and evokes a relevant behavior response. (A)
The apparatus: a magnesium head expander holding a 6-well dish mounted on a vertically oriented electrodynamic shaker housed inside a soundattenuation chamber. The stimulus consisted of a 60 Hz vertical displacement of the plate (hatched arrows) driven by an amplifier and controlled by a modified version of the Eaton-Peabody Laboratory Cochlear Function Test Suite. (B,D) Swimming behavior of 7-day-old larvae during exposure to the wave stimulus. Traces in (D) represent tracking of corresponding circled fish over 500 ms (1000 fps/ 500 frames). Asterisks indicate a "fast escape" response (B; inset). (C,E) Swimming behavior of larvae whose lateral-line neuromasts were ablated with low-dose CuSO4. Arrows in (E) indicate where a larva was swept into the waves and could no longer be tracked. Sisneros, 2013). To verify that the observed escape responses were mediated predominantly by 108 flow sensed by lateral-line hair cells rather than hair cells of the macula, we exposed a group of susceptible to disruption than L4, which was more susceptible to disruption than L3 (Fig. 2 F; exposure protocol that delivered intermittent pulses of stimulus ("periodic exposure"; Maximum intensity dorsal top-down 2D projections of confocal images of control or stimulus-exposed neuromast hair cells (blue (B) or orange (C); Parvalbumin immunolabel). Exposed neuromast hair-cell morphology was categorized as "normal" i.e. radial haircell organization indistinguishable from control or "disrupted" i.e. asymmetric organization with the hair-cell apical ends oriented posteriorly. (D) Maximum intensity projections of supporting cells (SCs) expressing GFP (green), immunolabeled synaptic ribbons (magenta; Ribeye b) and all cell nuclei (blue; DAPI). Note that SCs underlying displaced hair cells also appear physically disrupted (indicated by white arrows). Scale bars: 5 µm (F) Average percentage of neuromasts with "disrupted" morphology following mechanical stimulation. Each dot represents the percentage of disrupted neuromasts (NM) in a single experimental trial. Disrupted hair-cell morphology was place dependent, with neuromasts more frequently disrupted following sustained stimulus and when localized toward the posterior end of the tail (*P=0.0386, **P=0.0049, ***P=0.0004) (G) Average percentage of exposed neuromasts (NM) with "disrupted" morphology in lhfpl5b mutants, which lack mechanotransduction specifically in lateral-line hair cells, vs. heterozygous WT. lhfpl5b mutants show a similar gradient of neuromast disruption following mechanical injury as WT siblings. Error Bars = SEM test **P=0.0034), supporting that displacement of neuromasts is a consequence of mechanical 149 injury. Additionally, we examined hair cell morphology in the ears of larvae exposed to sustained 150 stimulus and observed no apparent damage (Supplemental Fig. 2), indicating our 151 overstimulation protocol produces mechanical damage specifically to lateral-line organs.

152
To determine if hair-cell activity plays a key role in the displacement of neuromasts, we 153 exposed lhfpl5b mutants-fish that have intact hair cell function in the ear, but no 154 mechanotransduction in hair cells of the lateral line-to sustained stimulation (Erickson,

155
Pacentine, Venuto, Clemens, & Nicolson, 2019). We observed comparable morphological 156 disruption of mutant neuromasts lacking mechanotransduction (Fig. 2 G), suggesting that showed "disrupted" hair-cell morphology immediately following sustained stimulus exposure. To 173 define the associations between overall neuromast morphology and specific structural changes 174 in mechanically injured neuromasts, we examined the numbers of hair cells and the percent of 175 hair cells contacted by afferent fibers in exposed neuromasts parsed into "normal" and 176 "disrupted" morphologies. With hair cell number, we observed significant loss specifically in 177 "disrupted" neuromasts, while "normal" neuromast hair-cell number appeared comparable to

Figure 3: Hair-cell loss and de-innervation is specific to "disrupted" neuromasts.
(A-C) Representative maximum intensity projection images of control (A) or exposed lateral line neuromasts with "normal" (B) or "disrupted" (C) morphology immediately following sustained strong wave exposure (0h). Synaptic ribbons (magenta; Ribeye b) and hair cells (blue; Parvalbumin) were immunolabled. Afferent neurons were expressing GFP. Scale bar: 5 µm (D) Hair-cell number per neuromast immediately post exposure. A significant reduction in hair cell number was observed (**Adj P=0.0019) and was specific to "disrupted" neuromasts (Adj P=0.3859 normal, ****Adj P<0.0001 disrupted). Pink box plot (Exp) represents pooled exposed neuromasts, while gray (Norm) and red (Dis) plots represent neuromasts parsed into normal and disrupted groups. Numbers beneath each plot indicate the number of neuromasts per group. Whiskers = min to max (E) Differences of least squares means in hair cell number per neuromast between groups. Bars represent 95% confidence interval (CI). (F) Percentage of neuromast hair cells innervated by afferent nerves. Numbers within each bar indicate the number of neuromasts per group. A significant portion of neuromast hair cells lacked afferent innervation following exposure (****Adj P<0.0001). Hair cells lacking afferent innervation were specifically observed in disrupted neuromasts (Adj P=0.7503 normal, ****Adj P<0.0001 disrupted). (G) Differences of least squares means in % hair cells innervated per neuromast between groups. Bars represent 95% CI In the larval zebrafish lateral line, afferent nerve fibers innervate multiple hair cells per 188 neuromast forming ~3-4 synaptic contacts per hair cell. To determine whether strong water 189 wave stimulus exposure generated lateral-line hair-cell synapse loss, we counted the number of 190 intact synapses (ribbons juxtaposed to PSDs; Fig. 4 A-C) in control and exposed larvae. We 191 observed significant reduction in the number of intact synapses per hair cell following sustained Figure 4: Significant hair-cell synapse loss is observed in "normal" neuromasts following mechanical overstimulation and exacerbated by blocking glutamate uptake. A-C) Representative maximum intensity projection images of unexposed (A), or stimulus exposed lateral-line neuromast with "normal" (B) or "disrupted" (C) morphology. Synaptic ribbons (magenta; Ribeye b), PSDs (green; MAGUK) and hair cells (blue; Parvalbumin) were immunolabled. Scale bar: 5 µm (D-E) Intact synapses per neuromast hair cell. Pink box plot in D (Exp) represents pooled exposed neuromasts while, in E, gray (Norm) and red (Dis) plots represent neuromasts parsed into normal and disrupted groups. Whiskers =min to max. The average number of intact synapses per hair cell was significantly reduced in exposed neuromasts (D; **Adj P=0.0078); when parsed, this reduction was significant in the "normal" exposure group relative to control (E; **Adj P=0.0043 normal, Adj P=0.1207 disrupted). (F) Differences of least squares means in number of intact synapses per hair cell between groups. Bars represent 95% CI. (G) The number of intact synapses per hair cell in larvae co-treated with TBOA, to block glutamate clearance, or drug carrier alone during exposure. Synapse loss was significantly greater in "normal" neuromasts coexposed to TBOA compared to fish co-exposed to the drug carrier alone (Two-way ANOVA. *P<0.0187). exposure ( Fig. 4 D, F; **Adj P=0.0078). When we 193 compared "normal" and "disrupted" neuromasts 194 following exposure, we observed a loss of intact 195 synapses per hair cell in all exposed neuromasts, 196 with significantly fewer synapses in "normal"   Representative images of control (A) and exposed (B) neuromasts. Synaptic ribbons (magenta; Ribeye b) and PSDs (green; MAGUK) were immunolabeled; hair cells were also immunolabeled, but not shown for clarity. Afferent neurons (white) were labeled with GFP. Insets: Arrows indicate intact synapses adjacent to afferent neurons; arrowheads (B) indicate synaptic debris. Scale bars: 5µm (main panels), 1µm (insets). (C) Frequency histogram of observed synaptic debris per neuromast (NM). While control neuromasts occasionally had 1 detached synapse, exposed neuromasts were observed that had up to 5 detached synapses. hair cells (Fig. 5 A,B). We then quantified instances where synapses, i.e. juxtaposed pre-and postsynaptic components associated with hair cells, were no longer adjacent to an afferent 228 nerve terminal. As these synapses appeared detached and suspended from their associated 229 neurons, making them no longer functional, we refer to them as synaptic debris. While we rarely 230 observed synaptic debris in unexposed neuromasts, we observed a greater relative frequency 231 of synaptic debris in neuromasts exposed to strong water wave stimulus (

248
which express YFP in all macrophages and microglia. Fish were fixed immediately after 249 exposure, or allowed to recover for 2, 4 or 8 hours. Control fish consisted of siblings that 250 received identical treatment but were not exposed to mechanical stimulation. Data were 251 obtained from the two terminal neuromasts from the pLL of each fish (Fig. 6 A). In agreement 252 with data shown in Fig. 3 D, we observed a modest but significant decline in hair-cell number in 253 specimens that were examined immediately and at 2 hours after sustained exposure (   suggesting that the inflammatory response was mediated by local macrophages and that 264 mechanical injury did not recruit macrophages from distant locations (Fig. 6 C). This pattern of 265 Figure 6. Macrophage response to mechanical overstimulation of lateral line hair cells. Experiments used fish that express YFP under regulation of the macrophage-specific mpeg1 promoter. All images and data were collected from the two distalmost neuromasts of the posterior lateral line (Fig 2A; term).

278
Recovery of hair-cell number occurred within 2-4 hours ( Fig. 6 A, B) and corresponded with 279 macrophages infiltrating neuromasts and phagocytosing hair-cell debris ( Fig. 6 A, E). We also 280 observed, compared to immediately following exposure, a lesser degree of afferent fiber

Figure 7: Mechanically overstimulated neuromasts recover hair-cell morphology, hair-cell number, and innervation. (A,B)
Average percentage of exposed neuromasts with "normal" vs. "disrupted" morphology following exposure. Each dot represents the percentage of disrupted neuromasts (L3-L5) in a single experimental trial; lines connect data points from the same cohort of exposed fish following 2 hours (A) or 48 hours (B) recovery. (C,D) Multilevel analysis of hair-cell number per neuromast immediately (0 h) post-exposure or after 2 or 48 hour recovery. Numbers beneath each plot indicate the number of neuromasts per group. Whiskers =min to max. Morphologically "disrupted" neuromasts has significantly fewer hair cells at 0h but not 2h following exposure (C; *Adj P=0.0321 (0h disrupted), Adj P=0.1875 (2h disrupted). Most exposed neuromasts were morphologically "normal" following 48h recovery and had a comparable number of hair cells relative to control (D; Adj P=0.4443). (E,F) The percentage of "disrupted" neuromast hair cells lacking afferent innervation was significant following 0h and 2h recovery (E; ****Adj P<0.0001 (0h disrupted), **Adj P=0.0016 (2h disrupted)). All hair cells were fully innervated following 48h recovery, including the few neuromasts with "disrupted" morphology (F; aligned red dots).
to ultrasonic waves and reported a delayed hair-cell death and synapse loss 48-72 hours 285 following exposure (Uribe et al., 2018). To determine if lateral line neuromasts exposed to the 286 strong wave stimulus generated by our apparatus underwent delayed hair-cell loss, we 287 examined hair-cell morphology, number, and innervation 48 hours following sustained stimulus 288 exposure. Most exposed neuromasts examined showed "normal" HC morphology (

Figure 8: Changes in synaptic ribbon and PSD sizes following sustained mechanical overstimulation. (A-A'')
Representative images of control (A) and exposed (A', A") neuromasts. Synaptic ribbons (magenta; Ribeye b), PSDs (green; MAGUK), and hair cells (blue, Parvalbumin) were immunolabeled. Scale bars: 5µm (main panels), 1µm (insets). (B-E) Box and whisker plots of relative synapse volumes normalized to 0h control. Whiskers indicate the min. and max. values; "+" indicates the mean value, horizontal lines indicate the relative median value of the control. (B) Ribbon volume appeared comparable to control immediately following exposure but was reduced 2 hours after exposure (*P=0.0195). (C) Significant reduction in ribbon size relative to control was specific to disrupted neuromasts (Kruskal-Wallis test: ***P=0.0004 (2h)). (D) Significantly larger PSDs were observed both immediately and 2 hours following exposure (****P<0.0001). (E) Enlarged PSDs were present in both "normal" and "disrupted" exposed neuromasts, with a greater enlargement observed 0h post-exposure (Kruskal-Wallis test: ****P<0.0001 (0h) ; ***P=0.0001, **P=0.0024 (2h)). postsynaptic components were also affected in our model, we compared the relative volumes of 296 neuromast hair-cell presynaptic ribbons and their corresponding PSDs in control and stimulus 297 exposed larvae. We observed a moderate reduction in synaptic-ribbon size following exposure; 298 ribbon volumes were significantly reduced relative to controls following 2 hours recovery ( Fig. 8 299 B; Kruskal-Wallis test *P=0.0195; N=3 trials), and this reduction was specific to "disrupted" 300 neuromasts (Fig. 8 C). While the changes in ribbon volume we observed were modest and 301 delayed in onset, we saw dramatic enlargement of PSDs immediately and 2 hours following  structures, we used confocal imaging and scanning electron microscopy (SEM) to assess hair 315 bundle morphology in both unexposed control larvae and larvae fixed immediately following 316 sustained exposure. All neuromasts throughout the fish were evaluated, but to remain 317 consistent with our fluorescence imaging results, we closely assessed the appearance of the 318 caudal pLL neuromasts. We found the caudal neuromasts to be more damaged than the ones 319 positioned more rostrally: the frequency of neuromasts with apparently disrupted appearance 320 increased the closer its position to the tail (Supplemental Fig. 3). This is consistent with our 321 fluorescence observations (Fig. 2 F; Supplemental Fig. 4 B) in which L5 neuromasts were more 322 likely to be disrupted than more anteriorly positioned L3.

323
A closer examination of neuromast morphology revealed a difference of the kinocilia after sustained exposure often appear to carry much shorter kinocilia ( Fig. 9 F-H, yellow 327 arrows), which lack bundling and, in some cases, pointing to different directions ( Fig. 9 G,H).

328
The apparent kinocilia length difference between control and overstimulated neuromasts 329 suggests at least some kinocilia may undergo a catastrophic damage event at the time of 330 stimulation, as their distal parts break off the hair cells. This is further supported by some 331

Figure 9: Scanning electron microscopy imaging of neuromasts following mechanical injury reveals disorganized hair cell stereocilia bundles and damaged kinocilia. (A-E)
Representative images of tail neuromasts of control fish larvae. Each hair cell carries a kinocilium, which is visibly thicker than its neighboring actin-filled, mechanosensitive stereocilia: see panel C featuring both structures at higher magnification (the kinocilium diameter is 220 nm, while stereocilia measured 90-110 nm). The kinocilia of control neuromasts are long (10-15 µm) and bundled together, while the stereocilia bundles have an apparent staircase arrangement. (F-K) Representative images of damaged tail neuromasts immediately following noise exposure featuring short (F-H, yellow arrows), disorganized (G, H), and swollen (I-K, yellow arrowheads) kinocilia, and disorganized stereocilia. (K) Same stereocilia bundle as in J marked with an asterisk at higher magnification to highlight the difference in the diameter of the kinocilium (360 nm) and neighboring stereocilia (85-100 nm) for noise exposed hair cells, as compared to the control hair cells in C. Average relative FM1-43FX fluorescence intensity measurements in control and exposed neuromasts over 4 hours of recovery. FM1-43FX uptake was significantly reduced in exposed neuromasts immediately following mechanical overstimulation but appear to completely recover by 4 hours (Tukey's multiple comparisons test ****P<0.0001 (0h), P=0.0579 (1h), P=0.8387 (2h), P=0.8387 (4h)). Dashed lines indicate FM1-43FX fluorescence intensity measurements in exposed neuromasts parsed into "normal" and "disrupted" morphologies. (E) Average stereocilia length of centrally localized hair bundles in control and exposed neuromasts.
Dashed lines indicate measurements in exposed neuromasts parsed into "normal" and "disrupted" morphologies. Error Bars = SD (F) Relative FM1-43FX fluorescence in both "normal" and "disrupted" exposed neuromasts was significantly reduced immediately following exposure but recovered over time (Tukey's multiple comparisons test ****P<0.0001, *P=0.0328 (0h); ****P<0.0001, **P=0.0098 (1h); **P=0.0025 control vs. dis, **P=0.0089 normal vs. dis (2h)). Each point represents an individual neuromast. Nearly all observed exposed neuromasts appeared morphologically normal following 4 hours; note only one neuromast data point in the disrupted category of the 4h recovery graph. Data were obtained from 26-32 neuromasts per condition over three trials. completely ( Fig. 9 I-K, yellow arrowheads). Accordingly, the average diameter of kinocilia at the 334 level of the hair bundle (L2-5) was significantly larger than control, with a few kinocilia showing 335 dramatically thicker widths ~2x greater than the thickest control (Fig. 9 L; ***P=0.0007). When 336 measured ~3-5 µm above the bundle, the exposed neuromast kinocilia have a somewhat larger 337 average diameter relative to control, but not as dramatic as observed at the base (Fig. 9 M; 338 *P=0.0243). Stereocilia bundles from both groups of animals carried tip links, but we were 339 unable to systematically evaluate and quantify their abundance. However, we observed signs of 340 damaged bundle morphology following overstimulation, as they often appeared splayed, with 341 gaps between the rows of stereocilia (Fig. 9 I-K).  348 10 A-C). We observed a significant reduction in the relative intensity of FM1-43 in all exposed 349 neuromasts immediately following exposure (Fig. 10 D, F (0h); ****P<0.0001). While phalloidin 350 labeling of stereocilia revealed what appeared to be tapered hair bundles in some exposed 351 neuromasts (Fig. 10 B'; yellow arrows), average stereocilia length obtained from 3D interpolated 352 confocal image stacks was not significantly altered (Fig. 10 E). Remarkably, FM1-43FX uptake Representative cross-section images of EdU (magenta) labeling of proliferating neuromast cells. Fish were exposed to EdU for 4 hours following stimulus exposure. Supporting cells (SC) were expressing GFP. Scale bars: 5µm (C) Average number of EdU+ cells per neuromast were comparable in control and exposed larvae. Data were obtained from 33-34 neuromasts per condition over three trials. (Two-way ANOVA. P=0.4193) Bars represent 95% CI.
showed recovery within 30 minutes and fully recovered over several hours. (Fig. 10 C, D). The 354 degree of FM1-43FX fluorescence recovery following mechanical damage appeared to 355 correspond with recovery of neuromast morphology; following 4 hours, nearly all neuromasts 356 exposed to strong wave stimulus showed "normal" morphology and relative FM1-43X 357 fluorescence that was comparable to control (Fig 10 D, F, 4h recovery). This observed timeline 358 of morphological recovery coincides with macrophage recruitment and phagocytosis hair cell 359 debris peaking 2 hours post exposure (Fig 6 D,E) followed by full recovery of hair cell number 360 between 2-4 hours post exposure (Fig 6 B). Additionally, we quantified proliferating neuromast

368
Using this method, we observed: i) hair-cell synapse loss in a subset of stimulus exposed   zebrafish and evoked a lateral-line mediated behavior (Fig 1 B, C) suggesting that the hair cells 397 were being directly stimulated by the water motion. The present method more closely resembles 398 the techniques that are typically used to study noise damage in the mammalian cochlea, where 399 high intensity acoustic energy causes hair cell and synaptic injury in specific regions of the 400 cochlea that are best-responsive to the frequency of the stimulus. This idea is further supported 401 by the observation that synapse loss in hair cells exposed to strong wave stimulation is greater 402 when glutamate uptake is blocked (Fig. 4 E), suggesting a shared mechanism of glutamate 403 excitotoxicity between noise-exposed mammalian ears and strong water wave stimulus exposed    In contrast to de-innervation and modest hair-cell loss we observed in mechanically 428 disrupted neuromasts, we saw significant loss of hair-cell synapses in neuromasts that were 429 exposed to strong water wave stimulus but not mechanically disrupted i.e. "normal" (Fig. 4). Two 430 notable observations were made in exposed neuromasts regarding synapse loss. First, loss of 431 synapses in "normal" exposed neuromasts was markedly more severe when synaptic glutamate 432 clearance was inhibited (Fig. 4 G), suggesting that synapse loss may reflect moderate hair-cell   439 3 B, Fig. 4 B). This observation was initially surprising given that pharmacologically activating 440 evolutionarily conserved Ca 2+ permeable AMPARs has been shown to drive afferent terminal 441 retraction in the zebrafish lateral line (Sebe et al., 2017). We propose that subtle damage to 442 afferents in "normal" exposed neuromasts may accompany synapse loss but not be apparent as 443 loss of innervation, as single afferent processes innervate multiple hair cells of the same polarity 444 within an individual neuromast (Dow, Jacobo, Hossain, Siletti, & Hudspeth, 2018; Faucherre, that the relative frequency of synaptic debris (i.e. synaptic components that appear detached 447 from afferent neurites) was higher in exposed neuromasts with "normal" morphology relative to 448 control (Fig. 5 C). We speculate that synaptic debris observed in exposed neuromasts with   mechanical overstimulation is much less than the injury that occurs after ototoxicity. We 470 observed macrophage entry in 30-40% of exposed neuromasts, despite modest hair cell loss 471 (Fig. 6 B,D). It is possible that the morphological changes characteristic of mechanically injured  Hair-cell synapse morphology following mechanical overstimulation 485 Immediately following mechanical overstimulation, the most pronounced morphological 486 change we observed in hair-cell synapses was significantly enlarged PSDs (Fig. 8 D, E).

518
We also found that mechanical trauma resulted in a small degree of hair-cell loss in 519 disrupted neuromasts (Fig. 3D), but that hair-cell numbers had recovered after 2 hours (

523
we observed a small amount of proliferation in neuromasts of both mechanically-damaged and 524 control fish. Since mechanical damage did not increase the level of cell proliferation in 525 neuromasts relative to control (Fig 11), we believe this observed cell division is likely associated 526 with the turnover process. Hair-cell regeneration in the vertebrate inner ear can also occur via 527 direct phenotypic conversion of supporting cells into a replacement hair cells (Warchol, 2011).

528
While it is conceivable that transdifferentiation of supporting cells could occur within 2 hours of 529 mechanical injury, such transdifferentiation has not been previously demonstrated in zebrafish 530 lateral line neuromasts (e.g. (Thomas et al., 2015)). Overall, these observations indicate that 531 mechanical trauma does not increase the rate of hair-cell production within neuromasts, further 532 supporting that neuromast recovery is largely due to hair-cell repair.

533
In summary, our data show that exposure of zebrafish lateral-line organs to strong water 534 wave results in mechanical injury and loss of afferent synapses, but that these injuries rapidly 535 recover. Our next steps will be to define the time course for synaptic recovery and to determine

613
Exposures consisted of 20 minutes of stimulation followed by a 10-minute break and 2 614 hours of uninterrupted stimulation. We also tested periodic exposures that consisted of a series 615 of short pulses spanning 2 hours total: 2 20-minute exposures each followed by 10 minutes of rest, followed by 30 minutes of stimulation, a 10-minute break, and a final 20 minutes of same conditions as exposed fish i.e. placed in a multi-well dish and maintained in the same room as the exposure chamber. For experiments pharmacologically blocking glutamate uptake, imaging, or allowed to recover for up to 2 days in an incubator at 29°C.

624
Free-swimming larvae were exposed to freshly made 3 µM CuSO4 solution in E3 for 1 625 hour, then rinsed and allowed to recover for 2 hours to ensure complete ablation of the lateral-626 line neuromasts. Neuromast ablation was confirmed by immunofluorescent labeling of hair cells.

627
The effects of low-dose copper exposure are likely specific to lateral-line organs; a previous 628 study in zebrafish determined exposure to low-dose CuSO4 for 1 hour did not alter the acoustic 629 escape response, which is similar to the fast start response we observed but evoked by higher     To selectively label hair-cell nuclei, live zebrafish larvae were incubated with DAPI 673 diluted 1:2000 in E3 media for 4 minutes. Larvae were briefly rinsed 3 times in fresh E3 media, 674 then immediately exposed to mechanical overstimulation.

755
To image hair-cell bundles, zebrafish larvae were exposed to strong water wave glutaraldehyde in 0.1 M sodium cacodylate buffer (Electron Microscopy Sciences) 758 supplemented with 2 mM CaCl2. Larvae were shipped overnight in fixative, then most of the 759 fixative (~90-95%) was removed, replaced with distilled water, and samples were stored at 4C.

768
(AIC) and Bayessian Information Criterion (BIC) were used to identify the best fitted covariance 769 structure. Tukey's adjustment was used for the alpha level to avoid type I error inflation due to 770 multiple comparisons. Estimated marginal mean differences and 95% Confidence Intervals 771 around them were explored and reported for quantification of effect size for group differences.

772
Graphs for data visualization and additional statistical analyses were performed Prism 8 773 (Graphpad Software Inc). Mixed model analysis was used to compare time-series data.

774
Statistical significance between synaptic ribbon and PSD volumes with was determined by 775 Kruskal-Wallis test (one independent variable) or Mann-Whitney U test (one independent 776 variable) and appropriate post-hoc tests. Based on the variance and effect sizes reported in 777 previous studies, the number of biological replicates were suitable to provide statistical power to Supplemental Figure 1: Fish exposed to periodic stimulus have less mechanical damage to neuromasts, but still show synapse loss. (A) Schematic of the two exposure protocols. Sustained exposure was a 20 min. pulse followed by 120 minutes uninterrupted mechanical overstimulation; periodic exposure was 90 min. exposure with intermittent 10 min. breaks totaling 120 minutes. (B) Periodic stimulus causes less neuromast disruption. Immediately following sustained exposure, 46% of exposed neuromasts showed a "disrupted" phenotype, whereas following a periodic exposure only 17% of the neuromasts appeared "disrupted" (Unpaired t-test **P=0.0034). (C) Position of the neuromast along the tail was also associated with vulnerability to disruption with both sustained and periodic stimulation.  Figure 2: Hair cell organs of the ear appeared undamaged in larvae exposed to sustained stimulus. Representative maximum intensity images of hair cell organs in the ears of control (A-C) and larvae exposed to sustained strong water wave stimulus (A'-C'). Hair cells in A-B were immunolabeled with an antibody against Otoferlin; posterior macula in C were immunolabled with antibodies against Parvalbumin to label hair cells and CtBP to label synaptic ribbons. Scale bars: 10 µm

Supplemental Figure 3: Scanning electron microscopy imaging of tail neuromasts following mechanical injury confirms the damage is more prominent for posterior neuromasts. (A-C)
Representative images of tail neuromasts of control fish larvae, presented as they are positioned on the larva: L1, L2 and terminal neuromasts. Each hair cell carries a tubulin-based primary cilium (kinocilium), which is thicker than the multiple mechanosensitive stereocilia arranged in a staircase. (D-G) Examples of tail neuromasts immediately following sustained stimulus exposure, presented as they are positioned on the larva: L2, L4, and two terminal neuromasts highlighting different levels of damage, with much more pronounced damage evident on terminal neuromasts. Scale bars: 1 µm

Supplemental Movies
Swimming behavior of 7-day-old larvae during exposure to the strong water wave stimulus shown in Figure 2 over 500ms (1000 fps/ 500 frames).
Movie S1: Swimming behavior of control fish with intact lateral line organs. Magenta circle indicates a fish prior to a"fast escape" response.
Movie S2: Swimming behavior of larvae whose lateral-line neuromasts were ablated with CuSO4. Magenta circles indicate larvae that were swept into the waves and could no longer tracked.