Our ability to perceive sounds relies on tiny cells deep inside our ears which can convert vibrations into the electrical signals that our brain is able to decode. These ‘hair cells’ sport a small tuft of short fibers on one of their ends that can move in response to pressure waves. The large amount of energy required for this activity is provided by the cells’ mitochondria, the small internal compartments that act as cellular powerhouses. In fact, reducing mitochondrial function in hair cells can lead to hearing disorders.

Mitochondria are often depicted as being bean-like, but they can actually adopt different shapes based on the level of energy they need to produce. Despite this link between morphology and function, little is known about what mitochondria look like in hair cells. Filling this knowledge gap is necessary to understand how these structures support hair cells and healthy hearing.

To address this question, McQuate et al. turned to zebrafish, as these animals detect vibrations in water through easily accessible hair cells on their skin that work just like the ones in the mammalian ear. Obtaining and analysing series of 3D images from a high-resolution microscope revealed that hair cells are more densely populated with mitochondria than other cell types. Mitochondrial organisation was also strikingly different. The side of the cell that carries the hair-like structures featured many small mitochondria; however, on the opposite side, which is in contact with neurons, the mitochondria formed a single large network. The co-existence of different types of mitochondria within one cell is a novel concept.

Further experiments investigated how these mitochondrial characteristics were connected to hair cell activity. They showed that this organisation was established gradually as the cells aged, with cellular activity shaping the architecture (but not the total volume) of the mitochondria.

Overall, the work by McQuate et al. provides important information necessary to develop therapeutics for hearing disorders linked to mitochondrial dysfunction. However, by showing that various kind of mitochondria can be present within one cell, it should also inform studies beyond those that focus on hearing.

Mitochondria are essential subcellular organelles in nearly all eukaryotic cells, where they perform and regulate manifold functions, including ATP production, calcium buffering, apoptosis, metabolite generation, among others. These functions are influenced by a cell’s total mitochondrial volume, regulated by mitochondrial biogenesis (Jornayvaz and Shulman, 2010) and subsequent mitochondrial architecture, sculpted by mitochondrial fusion and fission (Picard et al., 2013). Mitochondrial fusion and elongated mitochondria are associated with heightened mitochondrial membrane potentials, increased ATP production, and improved calcium buffering (Picard et al., 2013; Szabadkai et al., 2006; Gomes et al., 2011). Meanwhile, mitochondrial fission and smaller mitochondria are associated with lower mitochondrial membrane potentials, lower ATP production, and apoptosis (Liu et al., 2020). The combination of these features produces an overall mitochondrial phenotype according to cellular need. Failure to achieve an appropriate mitochondrial phenotype results in a variety of pathologies, particularly in highly metabolically active cells such as those that are electrically excitable (Reddy et al., 2011).

Hair cells (HCs) in the peripheral auditory nervous system mediate hearing and balance. Deflection of the stereocilia bundle at the apical pole of the HC results in cation influx in a process known as mechanotransduction, and subsequent depolarization and calcium influx through voltage-gated calcium channels (cav1.3) results in glutamate release from ribbon synapses at the basolateral pole onto afferent neurons. HCs heavily depend on mitochondria to sustain energetic demands, with 75% of their ATP usage produced via oxidative phosphorylation (Puschner and Schacht, 1997). It is perhaps due to their high dependency on mitochondria that HCs are particularly susceptible to mitochondrial alterations; mitochondria are implicated in both hereditary and environmentally induced hearing loss, as well as aging (Kokotas et al., 2007; Someya and Prolla, 2010; Böttger and Schacht, 2013). Mutations in over 30 mitochondrial-associated genes result in hearing loss in humans. These include mutations in the gene opa1, necessary for mitochondrial fusion (Leruez et al., 2013; Liguori et al., 2008).

Mitochondria are implicated at both poles of healthy HCs. In rat cochlear inner HCs, apical mitochondria have been shown to buffer calcium influx during mechanotransduction (Beurg et al., 2010; Pickett et al., 2018). Meanwhile, in zebrafish lateral line HCs, basal mitochondrial calcium uptake is essential for regulating ribbon size (Wong et al., 2019). Mitochondria also play a role in HC vulnerability to aminoglycoside exposure. HCs that have been treated with neomycin demonstrate abnormal mitochondrial morphologies prior to other insults (Owens et al., 2007). Neomycin-induced HC death requires mitochondrial calcium uptake (Esterberg et al., 2014 and Esterberg et al., 2016), and HC sensitivity to neomycin increases with cumulative mitochondrial activity (Pickett et al., 2018). Similarly, calcium import into the mitochondria via the MCU has been implicated in noise-induced hearing loss (Wang et al., 2018), and related mitochondrial potentials are disrupted in aging (Perkins et al., 2020).

Given the known impact of mitochondrial structure on their function, understanding the detailed morphological characteristics of HC mitochondria is a necessary first step for interpreting how these morphologies intersect with HC physiology and vulnerability (Lesus et al., 2019). We hypothesized that HCs maintained their mitochondria in an optimal configuration dependent on mitochondrial fusion to sustain high metabolic demands. Here, we detailed the characteristics of mitochondria in the HCs of the zebrafish lateral line. The lateral line is composed of clusters of HCs and surrounding supporting cells (SCs) called neuromasts (NMs) found on the surface of the fish’s body and detects changes in water flow. This information is vital for schooling and feeding behaviors. Lateral line HCs are genetically and morphologically similar to the HCs located within the mammalian inner ear, but are more easily accessible to genetic manipulations and experimentations (Pickett and Raible, 2019; Sheets et al., 2021). As fluorescent markers lack the resolution to distinguish between individual organelles, we turned to serial block-face scanning electron microscopy (SBFSEM) to produce three-dimensional reconstructions of HCs and their mitochondria with ultrastructural resolution. Over the course of this study, we reconstructed 5908 individual mitochondria from 162 different cells (16 NMs, 10 fish, Table 1). We demonstrated that zebrafish lateral line HCs had a mitochondrial phenotype distinct from SCs. This phenotype included a high total mitochondrial volume, and a particular architecture, with large, highly networked mitochondria at the basolateral pole near the synaptic ribbons, and smaller mitochondria positioned apically. This phenotype developed with HC maturation and was dependent on mechanotransduction and synaptic activity. Overall, our results demonstrate that HCs developed a highly specialized mitochondrial architecture sculpted by cell activity, which may explain their sensitivity to mitochondrial perturbation. Furthermore, the SBF datasets for each of the NMs used in this study have been made openly available, providing a comprehensive resource for further studying lateral line HCs and their organelles at the ultrastructural level.

Although mitochondria have long been known to be essential for HC function, a three-dimensional ultrastructural understanding of HC mitochondrial architecture, and its relationship to cell function, remained understudied. We used SBFSEM to describe an HC-specific mitochondrial phenotype characterized by (1) high mitochondrial volume and (2) a particular mitochondrial architecture, consisting of small mitochondria apically and large, networked mitochondria (max mito) near the synaptic ribbons. There are several caveats to using SBFSEM to study mitochondrial architecture. Although it is essential for resolving individual mitochondria, SBF requires fixed tissue, providing only a snapshot of the live mitochondrial dynamics that underlie the observed architecture. In addition, the laborious nature of the reconstructions precludes multiple replicates. However, we found no significant differences between HCs from different NMs and from different 5–6 dpf WT fish (65 HCs, 5 NMs from 3 fish), giving us confidence that our data reflect the greater population.

We found that lateral line HCs contain about 40 mitochondria in total. While this is over twice the number found in the surrounding SCs, it is notably on the smaller scale compared with numbers reported for other cell types, where estimates range from hundreds to thousands (Robin and Wong, 1988; Smith and Ord, 1983). We note these estimates were made from micrographs or measurements of mtDNA, which can vary widely. There are few studies that have similar complete reconstructions of total mitochondria in cells using high-resolution methodologies. A recent study indicates that there are nearly 500 mitochondria found in a primate cone photoreceptor after SBFSEM reconstruction (Hayes et al., 2021). Another recent study Liu et al., 2022 used SBFSEM to reconstruct mitochondria from mammalian cochlear HCs and estimated mitochondria numbered into the thousands. A better established measurement for comparison is mitochondrial volume as a fraction of the total cell volume (Posakony et al., 1977). Our measured ratio of mitochondrial volume to total cell volume in HCs (~7%, Figure 1H) is consistent with another recent study of mouse outer HCs that used EM tomography and estimated mitochondrial volume about 10% of cytoplasmic volume (Perkins et al., 2020). These numbers are also in alignment with SBFSEM reconstruction of neurons and glia in rat cortex, with mitochondrial volumes to cell volumes of 6–10% (Calì et al., 2019), though unfortunately mitochondrial number was not reported in this study.

We find distinct mitochondrial architecture along the HC apicobasal axis. At the HC apical pole, we find smaller mitochondria. In rat cochlear HCs, apical mitochondria take up calcium during mechanotransduction (Beurg et al., 2010), contributing to robust calcium buffering mechanics to maintain cytoplasmic calcium concentrations that could otherwise affect mechanotransduction and adaptation (Ricci et al., 1998; Eatock et al., 1987). Calcium influx itself could also lead to the smaller size of apical mitochondria. In neurons, calcium has been reported to induce mitochondrial fission (Rintoul et al., 2003), and the fission regulator Drp1 is activated by calcium (Cribbs and Strack, 2007; Han et al., 2008; Cereghetti et al., 2008). As excess calcium uptake can lead to mitochondrial damage (Starkov et al., 2004), mitochondrial fission in this region might facilitate mitophagy and quality control (Pfluger et al., 2015; Twig et al., 2008). The small size may also provide increased efficiency in packing, helping to maintain distinct calcium pools in apical and basal compartments, as has been suggested for photoreceptor mitochondria (Giarmarco et al., 2017). Mitochondria in opa1 mutants, which are uniformly small, demonstrated reduced calcium buffering. This could be a result of their small size or diminished mitochondrial health from lack of fusion, preventing resource sharing and hence leading to depolarized potentials. However, we note that mutations in opa1 will have multiple effects on mitochondrial function in addition to changes in size that may also influence their ability to take up calcium.

At the basolateral pole, HCs contained large mitochondrial networks (max mito) associated with ribbon synapses. Synaptic transmission requires large energetic expenditures (Li and Sheng, 2022). The proximity of the networked mitochondria to the synaptic ribbons suggests these large mitochondria might serve as active metabolic support. Mitochondrial fusion boosts oxidative phosphorylation and sharing of mitochondrial resources (Picard et al., 2013), indicating that the mitochondrial networks we observe would be particularly well-suited for this purpose. Synaptic mitochondria have been shown to take up calcium that enters through L-type CaV1.3 calcium channels, and disrupting this mitochondrial calcium uptake deregulated synaptic transmission (Wong et al., 2019). Larger mitochondria also have greater capacity to take up calcium (Kowaltowski et al., 2019; Szabadkai et al., 2006). As mitochondrial calcium uptake stimulates ATP production, interactions between mitochondrial fusion and synapse activity would be well-tuned to the cell’s energetic demands.

Our findings show that neither proper mechanotransduction nor synaptic activity is necessary for the growth in mitochondrial volume during HC maturation. This is surprising, given that HCs rely on oxidative phosphorylation for 75% of their metabolic needs (Puschner and Schacht, 1997), and changes in metabolic activity is a primary driver for mitochondrial biogenesis in many tissues (Jornayvaz and Shulman, 2010). Moreover, the role of calcium influx stimulating mitochondrial biogenesis in many other electrically active cell types, such as skeletal muscle and cardiomyocytes, is well established (Ojuka et al., 2002; Chin, 2004). We note that cdh23 mutations used to disrupt mechanotransduction still exhibit low levels of spontaneous synaptic release (Trapani and Nicolson, 2011) and cannot rule out the possibility that this residual activity might be sufficient to promote biogenesis. However, we can conclude that the growth of the mitochondrial volume is not regulated by overall activity levels. Additionally, we found that disrupting the mitochondrial architecture with a mutation in opa1 has no effect on the total mitochondrial volume. These fragmented mitochondria demonstrated depolarized potentials and decreased calcium uptake, likely associated with reduced ATP production. Together, these results suggest that the developmental increase in mitochondrial biogenesis is robust to changes in metabolic demands.

As the youngest HCs have a total mitochondrial volume similar to that of peripheral SCs, which serve as HC progenitors (Thomas and Raible, 2019), we suggest that the increase in mitochondrial volume is not linked to initial cell fate specification. Supporting this, the total mitochondrial volume increases gradually as HCs mature. This developmental mitochondrial growth is not driven by changes in cell volume, as described in other cell types (Rafelski et al., 2012; Miettinen and Björklund, 2017). Instead, it must be a product of different, ongoing pathways. One possibility may include the sirtuin deacetylases (SIRTs), key sensors of metabolism (Nogueiras et al., 2012) that can upregulate mitochondrial biogenesis through a series of parallel pathways (Yuan et al., 2016). In support of this, SIRT1 has been shown to be highly expressed in cochlear inner ear HCs (Xiong et al., 2014), and upregulation of SIRT1 pathways is protective against various modes of HC death (Zhan et al., 2021; Liang et al., 2021). The steady expansion of mitochondrial volume suggests these biogenesis-promoting pathways outweigh mitophagy and quality control. Such high mitochondrial volumes may produce dangerous reactive oxygen species (ROS) levels as a by-product of oxidative phosphorylation (Zhu et al., 2013). Therefore, the high mitochondrial volume might render HCs vulnerable to outside stresses that would additionally increase intracellular ROS. The observed mitochondrial networks might also in part counteract this vulnerability. Promoting mitochondrial fusion is protective against starvation-mediated apoptosis (Gomes et al., 2011) and SIRTs also regulate mitochondrial fusion (Uddin et al., 2021). SIRT3, located to mitochondria, has been implicated in counteracting age-related hearing loss and noise-induced damage (Someya and Prolla, 2010; Brown et al., 2014; Patel et al., 2020).

We found that mechanotransduction, while having little effect on mitochondrial biogenesis, was necessary for mitochondrial architecture. The clustered position of mechanotransduction mutants about midway within the mitochondrial UMAP trajectory (Figure 8) suggests a state of incomplete development where cells are unable to progress further. Lack of mechanotransduction resulted in fewer, larger, uniformly sized mitochondria. This is consistent with a model where apical calcium through mechanotransduction channels promotes mitochondrial fission. Mitochondria in mechanotransduction mutants also did not form basolateral networks localizing to synaptic ribbons. The fact that interference with mechanotransduction has effects on both formation of small mitochondria and localization of large, single networks suggests complex regulation of mitochondrial morphology.

By contrast, basal calcium entry associated with synaptic transmission is necessary for progressive growth of basal mitochondrial networks, but not formation of smaller apical mitochondria, which rapidly grew in number in cav1.3a mutants. In UMAP space, cav1.3a mutants intersperse with younger WT HCs, but segregate from mature HCs (Figure 8). Previous work (Trapani and Nicolson, 2011) has shown that cav1.3a mutants demonstrate reduced microphonic potentials, reflecting decreased mechanotransduction. If calcium influx drives apical mitochondrial fission, it is present at a level enough to do so in these mutants. Basal mitochondria take up calcium during synaptic transmission (Wong et al., 2019). This suggests a paradox where calcium might promote both mitochondrial fission apically and fusion basally. However, synaptic transmission requires large energy expenditures in addition to calcium regulation, and these demands may instead drive mitochondrial fusion into basal networks. Understanding the paradoxical effects of HC activity influencing both the formation of smaller apical mitochondria and larger basal networks will require additional study.

We find that altering calcium entry and synaptic transmission also altered ribbon size and morphology. These results compare with previous work showing that cav1.3a mutant HCs have enlarged ribbons (Sheets et al., 2012). Wong et al., 2019 demonstrated that blocking mitochondrial calcium entry also resulted in larger ribbons, a phenotype mimicked by altering NAD⁺/NADH ratios. While both of these studies showed larger ribbons after manipulating calcium and mitochondrial function, we observed cav1.3a mutant ribbons to be on average smaller than WT, but frequently found in clusters. We believe this discrepancy might be accounted for by the fact that the smaller ribbons we found in clusters at the EM level would appear to be larger single ribbons using fluorescence light microscopy methods employed in these previous studies. We also used strict structural parameters to define ribbons, which might account for differences in observed ribbon numbers and their locations as seen in fluorescence (Wong et al., 2019).

Not all HCs within zebrafish NMs are synaptically active, with active HCs tending to be younger cells on the NM periphery (Zhang et al., 2018; Wong et al., 2019; Lukasz et al., 2022). This is interesting, considering that we find the largest mitochondrial networks in more mature HCs. This discrepancy might be explained by considering that the initial onset of calcium entry associated with synaptic transmission might be sufficient to drive mitochondrial fusion, and that basolateral mitochondria remain networked regardless of whether the HC returns to a synaptically silent state during maturation.

Overall, our study provides a high-resolution, three-dimensional picture of zebrafish HC mitochondria, and demonstrates that through mechanotransduction and synaptic activity, these cells develop a finely tuned mitochondrial phenotype reflective of their function. This HC phenotype, through its high mitochondrial volume and large, appropriately positioned networked mitochondria, might be necessary to support high metabolic demands, but could also increase vulnerability to small fluctuations in intracellular ROS. Disruption of this phenotype could lead to improper HC physiology. Thus, it is critical that future studies take into account the high specificity with which HCs regulate their mitochondria.

The SBFSEM data volumes for the sixteen different NMs used in this manuscript, including WT and mutants, are publicly available at webKnossos (http://demo.webknossos.org/), searchable by the dataset name indicated in the relevant source data.

1. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4743,4995,263,0,1.3. 03052021_WT_left_2.

   https://webknossos.org/datasets/91969da369e1a848/03052021_WT_left_2/view#4743,4995,263,0,1.3
2. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4925,5724,246,0,1.3. 03052021_WT_SO1_Right.

   https://webknossos.org/datasets/91969da369e1a848/03052021_WT_SO1_Right/view#4925,5724,246,0,1.3
3. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4774,5185,341,0,1.3. 03052021_WT_SO2_Right.

   https://webknossos.org/datasets/91969da369e1a848/03052021_WT_SO2_Right/view#4774,5185,341,0,1.3
4. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4930,5395,267,0,1.3. 02102020_WT_6dpf_Right_3.

   https://webknossos.org/datasets/91969da369e1a848/02102020_WT_6dpf_Right_3/view#4930,5395,267,0,1.3
5. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4947,5529,222,0,1.3. 02102020_WT_6dpf.

   https://webknossos.org/datasets/91969da369e1a848/02102020_WT_6dpf/view#4947,5529,222,0,1.3
6. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4953,3852,212,0,1.3. 02102020_WT_3dpf_C.

   https://webknossos.org/datasets/91969da369e1a848/02102020_WT_3dpf_C/view#4953,3852,212,0,1.3
7. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 5297,5328,323,0,1.3. 02102020_WT_3dpfA_2.

   https://webknossos.org/datasets/91969da369e1a848/02102020_WT_3dpfA_2/view#5297,5328,323,0,1.3
8. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 3332,3547,205,0,1.3. 11262018_Opa1.

   https://webknossos.org/datasets/91969da369e1a848/11262018_Opa1/view#3332,3547,205,0,1.3
9. 1. McQuate AM
   2. Knecht S
   3. Raible D

   (2023) Webknossos

   ID 4623,4852,195,0,1.3. 12052019_cdh23_right.

   https://webknossos.org/datasets/91969da369e1a848/12052019_cdh23_right/view#4623,4852,195,0,1.3
10. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 5129,5047,160,0,1.3. 12052019_cdh23_left.

    https://webknossos.org/datasets/91969da369e1a848/12052019_cdh23_left/view#5129,5047,160,0,1.3
11. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 5690,5783,282,0,1.3. 12192019_cdh23_IO3.

    https://webknossos.org/datasets/91969da369e1a848/12192019_cdh23_IO3/view#5690,5783,282,0,1.3
12. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 5232,6593,288,0,1.3. 12192019_cdh23_fA_SO1.

    https://webknossos.org/datasets/91969da369e1a848/12192019_cdh23_fA_SO1/view#5232,6593,288,0,1.3
13. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 5932,6878,248,0,1.3. 07292020_CaV_A_Right.

    https://webknossos.org/datasets/91969da369e1a848/07292020_CaV_A_Right/view#5932,6878,248,0,1.3
14. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 6142,7029,322,0,1.3. 07292020_CaV_A_Left.

    https://webknossos.org/datasets/91969da369e1a848/07292020_CaV_A_Left/view#6142,7029,322,0,1.3
15. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID 4836,5799,367,0,1.3. 07292020_CaV_fishB_right.

    https://webknossos.org/datasets/91969da369e1a848/07292020_CaV_fishB_right/view#4836,5799,367,0,1.3
16. 1. McQuate AM
    2. Knecht S
    3. Raible D

    (2023) Webknossos

    ID webknossos.org. 07302020_cav_fishB_left_1.

    https://webknossos.org/datasets/91969da369e1a848/07302020_cav_fishB_left_1/view#5657,5364,352,0,1.3

1. 1. Beurg M
   2. Nam JH
   3. Chen Q
   4. Fettiplace R

   (2010) Calcium balance and mechanotransduction in rat cochlear hair cells

   Journal of Neurophysiology 104:18–34.

   https://doi.org/10.1152/jn.00019.2010

   • PubMed
   • Google Scholar
2. 1. Böttger EC
   2. Schacht J

   (2013) The mitochondrion: a perpetrator of acquired hearing loss

   Hearing Research 303:12–19.

   https://doi.org/10.1016/j.heares.2013.01.006

   • PubMed
   • Google Scholar
3. 1. Brown KD
   2. Maqsood S
   3. Huang JY
   4. Pan Y
   5. Harkcom W
   6. Li W
   7. Sauve A
   8. Verdin E
   9. Jaffrey SR

   (2014) Activation of SIRT3 by the NAD

   Cell Metabolism 20:1059–1068.

   https://doi.org/10.1016/j.cmet.2014.11.003

   • PubMed
   • Google Scholar
4. 1. Calì C
   2. Agus M
   3. Kare K
   4. Boges DJ
   5. Lehväslaiho H
   6. Hadwiger M
   7. Magistretti PJ

   (2019) 3D cellular reconstruction of cortical glia and parenchymal morphometric analysis from serial block-face electron microscopy of juvenile rat

   Progress in Neurobiology 183:101696.

   https://doi.org/10.1016/j.pneurobio.2019.101696

   • PubMed
   • Google Scholar
5. 1. Cardona A
   2. Saalfeld S
   3. Schindelin J
   4. Arganda-Carreras I
   5. Preibisch S
   6. Longair M
   7. Tomancak P
   8. Hartenstein V
   9. Douglas RJ

   (2012) TrakEM2 software for neural circuit reconstruction

   PLOS ONE 7:e38011.

   https://doi.org/10.1371/journal.pone.0038011

   • PubMed
   • Google Scholar
6. 1. Cereghetti GM
   2. Stangherlin A
   3. Martins de Brito O
   4. Chang CR
   5. Blackstone C
   6. Bernardi P
   7. Scorrano L

   (2008) Dephosphorylation by calcineurin regulates translocation of drp1 to mitochondria

   PNAS 105:15803–15808.

   https://doi.org/10.1073/pnas.0808249105

   • PubMed
   • Google Scholar
7. 1. Chin ER

   (2004) The role of calcium and calcium/calmodulin-dependent kinases in skeletal muscle plasticity and mitochondrial biogenesis

   The Proceedings of the Nutrition Society 63:279–286.

   https://doi.org/10.1079/PNS2004335

   • PubMed
   • Google Scholar
8. 1. Cribbs JT
   2. Strack S

   (2007) Reversible phosphorylation of Drp1 by cyclic AMP-dependent protein kinase and calcineurin regulates mitochondrial fission and cell death

   EMBO Reports 8:939–944.

   https://doi.org/10.1038/sj.embor.7401062

   • PubMed
   • Google Scholar
9. 1. Eatock RA
   2. Corey DP
   3. Hudspeth AJ

   (1987) Adaptation of mechanoelectrical transduction in hair cells of the bullfrog’s sacculus

   The Journal of Neuroscience 7:2821–2836.

   https://doi.org/10.1523/JNEUROSCI.07-09-02821.1987

   • PubMed
   • Google Scholar
10. 1. Esterberg R
    2. Hailey DW
    3. Rubel EW
    4. Raible DW

    (2014) Er-Mitochondrial calcium flow underlies vulnerability of mechanosensory hair cells to damage

    The Journal of Neuroscience 34:9703–9719.

    https://doi.org/10.1523/JNEUROSCI.0281-14.2014

    • PubMed
    • Google Scholar
11. 1. Esterberg R
    2. Linbo T
    3. Pickett SB
    4. Wu P
    5. Ou HC
    6. Rubel EW
    7. Raible DW

    (2016) Mitochondrial calcium uptake underlies ROS generation during aminoglycoside-induced hair cell death

    The Journal of Clinical Investigation 126:3556–3566.

    https://doi.org/10.1172/JCI84939

    • PubMed
    • Google Scholar
12. 1. Giarmarco MM
    2. Cleghorn WM
    3. Sloat SR
    4. Hurley JB
    5. Brockerhoff SE

    (2017) Mitochondria maintain distinct ca2+ pools in cone photoreceptors

    J Neurosci 37:2061–2072.

    https://doi.org/10.1523/JNEUROSCI.2689-16.2017

    • PubMed
    • Google Scholar
13. 1. Gomes LC
    2. Di Benedetto G
    3. Scorrano L

    (2011) During autophagy mitochondria elongate, are spared from degradation and sustain cell viability

    Nature Cell Biology 13:589–598.

    https://doi.org/10.1038/ncb2220

    • PubMed
    • Google Scholar
14. 1. Han XJ
    2. Lu YF
    3. Li SA
    4. Kaitsuka T
    5. Sato Y
    6. Tomizawa K
    7. Nairn AC
    8. Takei K
    9. Matsui H
    10. Matsushita M

    (2008) Cam kinase I alpha-induced phosphorylation of Drp1 regulates mitochondrial morphology

    The Journal of Cell Biology 182:573–585.

    https://doi.org/10.1083/jcb.200802164

    • PubMed
    • Google Scholar
15. 1. Hayes MJ
    2. Tracey-White D
    3. Kam JH
    4. Powner MB
    5. Jeffery G

    (2021) The 3D organisation of mitochondria in primate photoreceptors

    Sci Rep-Uk 11:18863.

    https://doi.org/10.1038/s41598-021-98409-7

    • Google Scholar
16. Book

    1. J. Deerinck T
    2. A. Bushong E
    3. H. Ellisman M
    4. Thor A

    (2022) Preparation of Biological Tissues for Serial Block Face Scanning Electron Microscopy SBEM V2

    University of California.

    https://doi.org/10.17504/protocols.io.36wgq7je5vk5/v2

    • Google Scholar
17. 1. Jornayvaz FR
    2. Shulman GI

    (2010) Regulation of mitochondrial biogenesis

    Essays Biochem 47:69–84.

    https://doi.org/10.1042/bse0470069

    • Google Scholar
18. 1. Kindt KS
    2. Finch G
    3. Nicolson T

    (2012) Kinocilia mediate mechanosensitivity in developing zebrafish hair cells

    Developmental Cell 23:329–341.

    https://doi.org/10.1016/j.devcel.2012.05.022

    • PubMed
    • Google Scholar
19. 1. Kokotas H
    2. Petersen MB
    3. Willems PJ

    (2007) Mitochondrial deafness

    Clinical Genetics 71:379–391.

    https://doi.org/10.1111/j.1399-0004.2007.00800.x

    • PubMed
    • Google Scholar
20. 1. Kowaltowski AJ
    2. Menezes-Filho SL
    3. Assali EA
    4. Gonçalves IG
    5. Cabral-Costa JV
    6. Abreu P
    7. Miller N
    8. Nolasco P
    9. Laurindo FRM
    10. Bruni-Cardoso A
    11. Shirihai OS

    (2019) Mitochondrial morphology regulates organellar ca 2+ uptake and changes cellular ca 2+ homeostasis

    FASEB J 33:13176–13188.

    https://doi.org/10.1096/fj.201901136r

    • PubMed
    • Google Scholar
21. 1. Leruez S
    2. Milea D
    3. Defoort-Dhellemmes S
    4. Colin E
    5. Crochet M
    6. Procaccio V
    7. Ferré M
    8. Lamblin J
    9. Drouin V
    10. Vincent-Delorme C
    11. Lenaers G
    12. Hamel C
    13. Blanchet C
    14. Juul G
    15. Larsen M
    16. Verny C
    17. Reynier P
    18. Amati-Bonneau P
    19. Bonneau D

    (2013) Sensorineural hearing loss in OPA1-linked disorders

    Brain 136:e236.

    https://doi.org/10.1093/brain/aws340

    • PubMed
    • Google Scholar
22. 1. Lesus J
    2. Arias K
    3. Kulaga J
    4. Sobkiv S
    5. Patel A
    6. Babu V
    7. Kambalyal A
    8. Patel M
    9. Padron F
    10. Mozaffari P
    11. Jayakumar A
    12. Ghatalah L
    13. Laban N
    14. Bahari R
    15. Perkins G
    16. Lysakowski A

    (2019) Why study inner ear hair cell mitochondria?

    HNO 67:429–433.

    https://doi.org/10.1007/s00106-019-0662-2

    • PubMed
    • Google Scholar
23. 1. Li S
    2. Sheng ZH

    (2022) Energy matters: presynaptic metabolism and the maintenance of synaptic transmission

    Nature Reviews. Neuroscience 23:4–22.

    https://doi.org/10.1038/s41583-021-00535-8

    • PubMed
    • Google Scholar
24. 1. Liang Z
    2. Zhang T
    3. Zhan T
    4. Cheng G
    5. Zhang W
    6. Jia H
    7. Yang H

    (2021) Metformin alleviates cisplatin-induced ototoxicity by autophagy induction possibly via the AMPK/FoxO3A pathway

    Journal of Neurophysiology 125:1202–1212.

    https://doi.org/10.1152/jn.00417.2020

    • PubMed
    • Google Scholar
25. 1. Liguori M
    2. La Russa A
    3. Manna I
    4. Andreoli V
    5. Caracciolo M
    6. Spadafora P
    7. Cittadella R
    8. Quattrone A

    (2008) A phenotypic variation of dominant optic atrophy and deafness (ADOAD) due to a novel OPA1 mutation

    Journal of Neurology 255:127–129.

    https://doi.org/10.1007/s00415-008-0571-x

    • PubMed
    • Google Scholar
26. 1. Liu YJ
    2. McIntyre RL
    3. Janssens GE
    4. Houtkooper RH

    (2020) Mitochondrial fission and fusion: a dynamic role in aging and potential target for age-related disease

    Mechanisms of Ageing and Development 186:111212.

    https://doi.org/10.1016/j.mad.2020.111212

    • PubMed
    • Google Scholar
27. 1. Liu J
    2. Wang S
    3. Lu Y
    4. Wang H
    5. Wang F
    6. Qiu M
    7. Xie Q
    8. Han H
    9. Hua Y

    (2022) Aligned organization of synapses and mitochondria in auditory hair cells

    Neuroscience Bulletin 38:235–248.

    https://doi.org/10.1007/s12264-021-00801-w

    • PubMed
    • Google Scholar
28. 1. Lukasz D
    2. Beirl A
    3. Kindt K

    (2022) Chronic neurotransmission increases the susceptibility of lateral-line hair cells to ototoxic insults

    eLife 11:e77775.

    https://doi.org/10.7554/eLife.77775

    • PubMed
    • Google Scholar
29. 1. McInnes L
    2. Healy J
    3. Saul N
    4. Großberger L

    (2018) UMAP: uniform manifold approximation and projection

    Journal of Open Source Software 3:861.

    https://doi.org/10.21105/joss.00861

    • Google Scholar
30. 1. Miettinen TP
    2. Björklund M

    (2017) Mitochondrial function and cell size: an allometric relationship

    Trends in Cell Biology 27:393–402.

    https://doi.org/10.1016/j.tcb.2017.02.006

    • PubMed
    • Google Scholar
31. 1. Nogueiras R
    2. Habegger KM
    3. Chaudhary N
    4. Finan B
    5. Banks AS
    6. Dietrich MO
    7. Horvath TL
    8. Sinclair DA
    9. Pfluger PT
    10. Tschöp MH

    (2012) Sirtuin 1 and sirtuin 3: physiological modulators of metabolism

    Physiological Reviews 92:1479–1514.

    https://doi.org/10.1152/physrev.00022.2011

    • PubMed
    • Google Scholar
32. 1. Ojuka EO
    2. Jones TE
    3. Han DH
    4. Chen M
    5. Wamhoff BR
    6. Sturek M
    7. Holloszy JO

    (2002) Intermittent increases in cytosolic ca2+ stimulate mitochondrial biogenesis in muscle cells

    American Journal of Physiology. Endocrinology and Metabolism 283:E1040–E1045.

    https://doi.org/10.1152/ajpendo.00242.2002

    • PubMed
    • Google Scholar
33. 1. Olichon A
    2. Guillou E
    3. Delettre C
    4. Landes T
    5. Arnauné-Pelloquin L
    6. Emorine LJ
    7. Mils V
    8. Daloyau M
    9. Hamel C
    10. Amati-Bonneau P
    11. Bonneau D
    12. Reynier P
    13. Lenaers G
    14. Belenguer P

    (2006) Mitochondrial dynamics and disease, OPA1

    Biochimica et Biophysica Acta 1763:500–509.

    https://doi.org/10.1016/j.bbamcr.2006.04.003

    • PubMed
    • Google Scholar
34. 1. Olt J
    2. Johnson SL
    3. Marcotti W

    (2014) In vivo and in vitro biophysical properties of hair cells from the lateral line and inner ear of developing and adult zebrafish

    The Journal of Physiology 592:2041–2058.

    https://doi.org/10.1113/jphysiol.2013.265108

    • PubMed
    • Google Scholar
35. 1. Owens KN
    2. Cunningham DE
    3. MacDonald G
    4. Rubel EW
    5. Raible DW
    6. Pujol R

    (2007) Ultrastructural analysis of aminoglycoside-induced hair cell death in the zebrafish lateral line reveals an early mitochondrial response

    The Journal of Comparative Neurology 502:522–543.

    https://doi.org/10.1002/cne.21345

    • PubMed
    • Google Scholar
36. 1. Patel S
    2. Shah L
    3. Dang N
    4. Tan X
    5. Almudevar A
    6. White PM

    (2020) Sirt3 promotes auditory function in young adult FVB/nj mice but is dispensable for hearing recovery after noise exposure

    PLOS ONE 15:e0235491.

    https://doi.org/10.1371/journal.pone.0235491

    • PubMed
    • Google Scholar
37. Preprint

    1. Pedregosa F
    2. Varoquaux G
    3. Gramfort A
    4. Michel V
    5. Thirion B
    6. Grisel O
    7. Blondel M
    8. Müller A
    9. Nothman J
    10. Louppe G
    11. Prettenhofer P
    12. Weiss R
    13. Dubourg V
    14. Vanderplas J
    15. Passos A
    16. Cournapeau D
    17. Brucher M
    18. Perrot M
    19. Duchesnay É

    (2012) Scikit-Learn: Machine Learning in Python

    arXiv.

    https://doi.org/10.48550/arxiv.1201.0490

    • Google Scholar
38. 1. Perkins GA
    2. Tjong J
    3. Brown JM
    4. Poquiz PH
    5. Scott RT
    6. Kolson DR
    7. Ellisman MH
    8. Spirou GA

    (2010) The micro-architecture of mitochondria at active zones: electron tomography reveals novel anchoring scaffolds and cristae structured for high-rate metabolism

    The Journal of Neuroscience 30:1015–1026.

    https://doi.org/10.1523/JNEUROSCI.1517-09.2010

    • PubMed
    • Google Scholar
39. 1. Perkins G
    2. Lee JH
    3. Park S
    4. Kang M
    5. Perez-Flores MC
    6. Ju S
    7. Phillips G
    8. Lysakowski A
    9. Gratton MA
    10. Yamoah EN

    (2020) Altered outer hair cell mitochondrial and subsurface cisternae connectomics are candidate mechanisms for hearing loss in mice

    The Journal of Neuroscience 40:8556–8572.

    https://doi.org/10.1523/JNEUROSCI.2901-19.2020

    • PubMed
    • Google Scholar
40. 1. Pfluger PT
    2. Kabra DG
    3. Aichler M
    4. Schriever SC
    5. Pfuhlmann K
    6. García VC
    7. Lehti M
    8. Weber J
    9. Kutschke M
    10. Rozman J
    11. Elrod JW
    12. Hevener AL
    13. Feuchtinger A
    14. Hrabě de Angelis M
    15. Walch A
    16. Rollmann SM
    17. Aronow BJ
    18. Müller TD
    19. Perez-Tilve D
    20. Jastroch M
    21. De Luca M
    22. Molkentin JD
    23. Tschöp MH

    (2015) Calcineurin links mitochondrial elongation with energy metabolism

    Cell Metabolism 22:838–850.

    https://doi.org/10.1016/j.cmet.2015.08.022

    • PubMed
    • Google Scholar
41. 1. Picard M
    2. Shirihai OS
    3. Gentil BJ
    4. Burelle Y

    (2013) Mitochondrial morphology transitions and functions: implications for retrograde signaling?

    American Journal of Physiology. Regulatory, Integrative and Comparative Physiology 304:R393–R406.

    https://doi.org/10.1152/ajpregu.00584.2012

    • PubMed
    • Google Scholar
42. 1. Pickett SB
    2. Thomas ED
    3. Sebe JY
    4. Linbo T
    5. Esterberg R
    6. Hailey DW
    7. Raible DW

    (2018) Cumulative mitochondrial activity correlates with ototoxin susceptibility in zebrafish mechanosensory hair cells

    eLife 7:e38062.

    https://doi.org/10.7554/eLife.38062

    • PubMed
    • Google Scholar
43. 1. Pickett SB
    2. Raible DW

    (2019) Water waves to sound waves: using zebrafish to explore hair cell biology

    Journal of the Association for Research in Otolaryngology 20:1–19.

    https://doi.org/10.1007/s10162-018-00711-1

    • PubMed
    • Google Scholar
44. 1. Posakony JW
    2. England JM
    3. Attardi G

    (1977) Mitochondrial growth and division during the cell cycle in HeLa cells

    The Journal of Cell Biology 74:468–491.

    https://doi.org/10.1083/jcb.74.2.468

    • PubMed
    • Google Scholar
45. 1. Puschner B
    2. Schacht J

    (1997) Energy metabolism in cochlear outer hair cells in vitro

    Hearing Research 114:102–106.

    https://doi.org/10.1016/s0378-5955(97)00163-9

    • PubMed
    • Google Scholar
46. 1. Rafelski SM
    2. Viana MP
    3. Zhang Y
    4. Chan YHM
    5. Thorn KS
    6. Yam P
    7. Fung JC
    8. Li H
    9. Costa L
    10. Marshall WF

    (2012) Mitochondrial network size scaling in budding yeast

    Science 338:822–824.

    https://doi.org/10.1126/science.1225720

    • PubMed
    • Google Scholar
47. 1. Raible DW
    2. Kruse GJ

    (2000) Organization of the lateral line system in embryonic zebrafish

    The Journal of Comparative Neurology 421:189–198.

    https://doi.org/10.1002/(sici)1096-9861(20000529)421:23.0.co;2-k

    • PubMed
    • Google Scholar
48. 1. Reddy PH
    2. Reddy TP
    3. Manczak M
    4. Calkins MJ
    5. Shirendeb U
    6. Mao P

    (2011) Dynamin-Related protein 1 and mitochondrial fragmentation in neurodegenerative diseases

    Brain Research Reviews 67:103–118.

    https://doi.org/10.1016/j.brainresrev.2010.11.004

    • PubMed
    • Google Scholar
49. 1. Rey SJ
    2. Anselin L

    (2007) PySAL: a python library of spatial analytical methods

    Review of Regional Studies 37:8285.

    https://doi.org/10.52324/001c.8285

    • PubMed
    • Google Scholar
50. 1. Ricci AJ
    2. Wu YC
    3. Fettiplace R

    (1998) The endogenous calcium buffer and the time course of transducer adaptation in auditory hair cells

    The Journal of Neuroscience 18:8261–8277.

    https://doi.org/10.1523/JNEUROSCI.18-20-08261.1998

    • PubMed
    • Google Scholar
51. 1. Rintoul GL
    2. Filiano AJ
    3. Brocard JB
    4. Kress GJ
    5. Reynolds IJ

    (2003) Glutamate decreases mitochondrial size and movement in primary forebrain neurons

    The Journal of Neuroscience 23:7881–7888.

    https://doi.org/10.1523/JNEUROSCI.23-21-07881.2003

    • PubMed
    • Google Scholar
52. 1. Robin ED
    2. Wong R

    (1988) Mitochondrial DNA molecules and virtual number of mitochondria per cell in mammalian cells

    Journal of Cellular Physiology 136:507–513.

    https://doi.org/10.1002/jcp.1041360316

    • PubMed
    • Google Scholar
53. 1. Romero-Carvajal A
    2. Navajas Acedo J
    3. Jiang L
    4. Kozlovskaja-Gumbrienė A
    5. Alexander R
    6. Li H
    7. Piotrowski T

    (2015) Regeneration of sensory hair cells requires localized interactions between the Notch and Wnt pathways

    Developmental Cell 34:267–282.

    https://doi.org/10.1016/j.devcel.2015.05.025

    • PubMed
    • Google Scholar
54. 1. Shah AN
    2. Davey CF
    3. Whitebirch AC
    4. Miller AC
    5. Moens CB

    (2015) Rapid reverse genetic screening using CRISPR in zebrafish

    Nature Methods 12:535–540.

    https://doi.org/10.1038/nmeth.3360

    • PubMed
    • Google Scholar
55. 1. Sheets L
    2. Kindt KS
    3. Nicolson T

    (2012) Presynaptic Cav1.3 channels regulate synaptic ribbon size and are required for synaptic maintenance in sensory hair cells

    The Journal of Neuroscience 32:17273–17286.

    https://doi.org/10.1523/JNEUROSCI.3005-12.2012

    • PubMed
    • Google Scholar
56. 1. Sheets L
    2. Holmgren M
    3. Kindt KS

    (2021) How zebrafish can drive the future of genetic-based hearing and balance research

    Journal of the Association for Research in Otolaryngology 22:215–235.

    https://doi.org/10.1007/s10162-021-00798-z

    • PubMed
    • Google Scholar
57. 1. Smith RA
    2. Ord MJ

    (1983) Mitochondrial form and function relationships in vivo: their potential in toxicology and pathology

    International Review of Cytology 83:63–134.

    https://doi.org/10.1016/s0074-7696(08)61686-1

    • PubMed
    • Google Scholar
58. 1. Söllner C
    2. Rauch G-J
    3. Siemens J
    4. Geisler R
    5. Schuster SC
    6. Müller U
    7. Nicolson T
    8. Tübingen 2000 Screen Consortium

    (2004) Mutations in cadherin 23 affect tip links in zebrafish sensory hair cells

    Nature 428:955–959.

    https://doi.org/10.1038/nature02484

    • PubMed
    • Google Scholar
59. 1. Someya S
    2. Prolla TA

    (2010) Mitochondrial oxidative damage and apoptosis in age-related hearing loss

    Mechanisms of Ageing and Development 131:480–486.

    https://doi.org/10.1016/j.mad.2010.04.006

    • PubMed
    • Google Scholar
60. 1. Starkov AA
    2. Chinopoulos C
    3. Fiskum G

    (2004) Mitochondrial calcium and oxidative stress as mediators of ischemic brain injury

    Cell Calcium 36:257–264.

    https://doi.org/10.1016/j.ceca.2004.02.012

    • PubMed
    • Google Scholar
61. 1. Stowers RS
    2. Megeath LJ
    3. Górska-Andrzejak J
    4. Meinertzhagen IA
    5. Schwarz TL

    (2002) Axonal transport of mitochondria to synapses depends on Milton, a novel Drosophila protein

    Neuron 36:1063–1077.

    https://doi.org/10.1016/s0896-6273(02)01094-2

    • PubMed
    • Google Scholar
62. 1. Szabadkai G
    2. Simoni AM
    3. Bianchi K
    4. De Stefani D
    5. Leo S
    6. Wieckowski MR
    7. Rizzuto R

    (2006) Mitochondrial dynamics and Ca2+ signaling

    Biochimica et Biophysica Acta 1763:442–449.

    https://doi.org/10.1016/j.bbamcr.2006.04.002

    • PubMed
    • Google Scholar
63. 1. Thomas ED
    2. Raible DW

    (2019) Distinct progenitor populations mediate regeneration in the zebrafish lateral line

    eLife 8:e43736.

    https://doi.org/10.7554/eLife.43736

    • PubMed
    • Google Scholar
64. 1. Trapani JG
    2. Nicolson T

    (2011) Mechanism of spontaneous activity in afferent neurons of the zebrafish lateral-line organ

    The Journal of Neuroscience 31:1614–1623.

    https://doi.org/10.1523/JNEUROSCI.3369-10.2011

    • PubMed
    • Google Scholar
65. 1. Twig G
    2. Hyde B
    3. Shirihai OS

    (2008) Mitochondrial fusion, fission and autophagy as a quality control axis: the bioenergetic view

    Biochimica et Biophysica Acta 1777:1092–1097.

    https://doi.org/10.1016/j.bbabio.2008.05.001

    • PubMed
    • Google Scholar
66. 1. Uddin GM
    2. Abbas R
    3. Shutt TE

    (2021) The role of protein acetylation in regulating mitochondrial fusion and fission

    Biochemical Society Transactions 49:2807–2819.

    https://doi.org/10.1042/BST20210798

    • PubMed
    • Google Scholar
67. 1. Wang X
    2. Zhu Y
    3. Long H
    4. Pan S
    5. Xiong H
    6. Fang Q
    7. Hill K
    8. Lai R
    9. Yuan H
    10. Sha SH

    (2018) Mitochondrial calcium transporters mediate sensitivity to noise-induced losses of hair cells and cochlear synapses

    Frontiers in Molecular Neuroscience 11:469.

    https://doi.org/10.3389/fnmol.2018.00469

    • PubMed
    • Google Scholar
68. 1. Wong H-TC
    2. Zhang Q
    3. Beirl AJ
    4. Petralia RS
    5. Wang Y-X
    6. Kindt K

    (2019) Synaptic mitochondria regulate hair-cell synapse size and function

    eLife 8:e48914.

    https://doi.org/10.7554/eLife.48914

    • PubMed
    • Google Scholar
69. 1. Xiong H
    2. Dai M
    3. Ou Y
    4. Pang J
    5. Yang H
    6. Huang Q
    7. Chen S
    8. Zhang Z
    9. Xu Y
    10. Cai Y
    11. Liang M
    12. Zhang X
    13. Lai L
    14. Zheng Y

    (2014) Sirt1 expression in the cochlea and auditory cortex of a mouse model of age-related hearing loss

    Experimental Gerontology 51:8–14.

    https://doi.org/10.1016/j.exger.2013.12.006

    • PubMed
    • Google Scholar
70. 1. Yuan Y
    2. Cruzat VF
    3. Newsholme P
    4. Cheng J
    5. Chen Y
    6. Lu Y

    (2016) Regulation of SIRT1 in aging: roles in mitochondrial function and biogenesis

    Mechanisms of Ageing and Development 155:10–21.

    https://doi.org/10.1016/j.mad.2016.02.003

    • PubMed
    • Google Scholar
71. 1. Zhan T
    2. Xiong H
    3. Pang J
    4. Zhang W
    5. Ye Y
    6. Liang Z
    7. Huang X
    8. He F
    9. Jian B
    10. He W
    11. Gao Y
    12. Min X
    13. Zheng Y
    14. Yang H

    (2021) Modulation of NAD+ biosynthesis activates SIRT1 and resists cisplatin-induced ototoxicity

    Toxicology Letters 349:115–123.

    https://doi.org/10.1016/j.toxlet.2021.05.013

    • PubMed
    • Google Scholar
72. 1. Zhang Q
    2. Li S
    3. Wong HTC
    4. He XJ
    5. Beirl A
    6. Petralia RS
    7. Wang YX
    8. Kindt KS

    (2018) Synaptically silent sensory hair cells in zebrafish are recruited after damage

    Nature Communications 9:1388.

    https://doi.org/10.1038/s41467-018-03806-8

    • PubMed
    • Google Scholar
73. 1. Zhu J
    2. Wang KZ
    3. Chu CT

    (2013) After the banquet

    Autophagy 9:1663–1676.

    https://doi.org/10.4161/auto.24135

    • Google Scholar

Decision letter after peer review:

Thank you for submitting your article "Activity regulates a cell type-specific mitochondrial phenotype in zebrafish lateral line hair cells" for consideration by eLife. Your article has been reviewed by 3 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Didier Stainier as the Senior Editor. The following individual involved in the review of your submission has agreed to reveal their identity: Katie Kindt (Reviewer #1).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission.

Essential revisions:

All three reviewers thought this high-resolution mitochondrial study in sensory hair cells is novel and interesting. Nevertheless, there are some valid concerns raised by the reviewers that should be improved such as:

1) A better link between mitochondrial disruption to hair cell function in opa1 mutants,

2) Improve data analyses, and

3) Better functional evidence to support the hypothesis based on morphological analysis.

We will be happy to look at a revision of this manuscript should you be able to address the reviewers' concerns.

Reviewer #1 (Recommendations for the authors):

This is a rich dataset and resource for the community that has been thoroughly and thoughtfully analyzed. The presentation of the data and figures is high quality and thoughtfully executed. Our comments seek to further strengthen the manuscript.

1. Please provide a table(s) to summarize the EM sample data, including the age, neuromast identity/identities, genotype, whether the samples are whole or partial neuromasts, which cells have been reconstructed (HCs, SCs), and a list of the figures that use data from that sample.

2. Include information in your introduction that highlights the total number of samples reconstructed and the public availability of the data. This is an amazing resource for the community and is worth highlighting more.

3. For cdh23 and cav1.3 mutants, please provide a comparison of wt to mutant mitochondrial volume/total cell volume as in Figure S1B. It looks like the HC cell size may be different in some of the mutants, which may impact mitochondrial volume.

4. Sample sizes. How many different fish were analyzed for P-SCs? The number of NMs and fish are different in different figure legends. If n = 1 fish, there should be more than 1 NM and more than 1 fish analyzed for P-SC mitos. It would seem that the data already exists (as there are 2+ fish for both HCs and C-SCs), so it should be possible to at least increase these numbers to 2 NMs and 2 fish for all cell types. More broadly, multiple fish would ideally be examined for each mutant. We recognize that this is incredibly labor-intensive work, but recent EM datasets published in eLife have shown data for n>1 even in mutants (ex. Dow et al. 2018).

5. Can you show how the position of the max mito changes over developmental time for Figures 3 and 4?

6. In a supplement it would be helpful to add a plot showing the sizes of all mitochondria for a mature, representative HC(s)? It would be nice to see an example(s) of the full distribution.

7. Opa1 mutations have a broad range of effects, as mitochondrial fusion is hypothesized to impact many aspects of mitochondrial function. This makes it hard to link functional changes in these mutants to changes in mito architecture. Please discuss this caveat.

8. Although evoked mitoGCaMP3 responses in opa-1 mutants are compelling, changes in evoked responses could also be due to defects in mechanotransduction. Another method to verify normal mechanotransduction would be helpful. For example, normal FM 1-43 labeling or unaltered evoked cyto- or mem-GCaMP6 responses.

Reviewer #2 (Recommendations for the authors):

Results and Figures:

– Labeling of larval age and hair cell age and position is inconsistent throughout the results and figures. For example, in Figure 2 A neither larval age nor hair cell position are specified.

– Line 139-140: "Younger hair cells on periphery and mature hair cells in center" need a citation.

– Development of mitochondria with maturation (Figure 3): the authors show images of representative neuromasts from younger (3 dpf) and more mature (5-6 dpf) neuromasts in Figure 3A, yet analyzed pooled mitochondrial morphologies from all three ages using the position of the hair cell in the neuromast as a proxy for maturity. Given that 3 dpf neuromasts are not yet functionally mature and do not show concentric patterning comparable to older neuromasts, it is unclear why the authors used this approach. The analysis shown in Figure 3B-E would be more convincing if central and peripheral hair cell data were shown for just 5-6 dpf, with 3 dpf shown separately.

– Figure 4A: How were young, intermediate, and mature hair cells defined?

– Figure 5D-F: Is this data pooled from all 3 ages?

– UMAP data/ Figure 6 and Figure 10: A more detailed description of UMAP data analysis and what it specifically reveals about mitochondrial properties in relation to hair cell maturation and/or function is needed.

– opa1 mutants/ Figure 7: The authors show reduced evoked mitochondrial calcium uptake in opa1 mutants. Did they examine mitoGCaMP levels at rest (i.e. compare mitoGCaMP fluorescence in WT vs mutant without stimulation)? Additionally, there is a possibility that the hair cells in opa1 mutants have impaired mechanotransduction. Did the authors examine FM1-43 uptake in mutants vs. WT?

– cdh23 mutants/ Figure 8: To support the conclusion that the largest mitochondria do not localize to the presynaptic ribbons, the authors should show the localization of the ribbons in the representative reconstructions in Figure 8A.

– cav1.3a mutants/ Figure 9: This analysis overall would be far more impactful if it corresponded with the analysis performed on cdh23 mutants. The authors should repeat the analysis of the data in the same way it was performed for cdh23 mutants, then expand on the cav1.3a mutant analysis with an additional figure.

– UMAP data/ Figure 10: In Figure 10A, does the WT condition contain cells from all ages (3 and 5/6 dpf)? If so, that additional variable may be confounding, as the two mutants examined only represent hair cells from 5 dpf fish.

Discussion:

– The discussion on the possible mechanisms for distinct mitochondrial morphology at the apical and basolateral ends of hair cells needs some revising. It's unclear to this reviewer how the data shown supports that calcium signaling shapes mitochondrial morphology through fission. The authors propose that calcium influx through MET may itself contribute to smaller mitochondria, yet large amounts of calcium enter presynaptic Cav1.3a channels when active hair cells are stimulated, and this is where the mitochondrial volume is the largest. Conversely, while the authors suggest that large mitochondrial networks near the synapses may be specifically necessary to drive PMCA1 pumps, it should be noted that PMCA2 pumps play a critical role in the regulation of calcium at hair cell stereocilia. Overall, it seems that other key differences, such as the high metabolic demand of synaptic transmission itself, may play a role in the localization of large mitochondria to synapses.

Materials and methods:

– More details are needed for waterjet and calcium imaging. What concentration of MESAB was used? What was the composition of the extracellular solution? Which neuromast was acquired?

Reviewer #3 (Recommendations for the authors):

1. Some data from the too small sample size. In particular, P-SC in Figure 1 and 2, and Opa1 mutant in Figure 6A, and B are from only 1 fish. Although the authors pointed out the laborious nature of the reconstructions and suggested that their data reflect the greater population in the Discussion section, the reviewer believes that their reconstruction method and results are the central topics of this study. Therefore, the description of too much difficulty and speculation will make their promising data and techniques unattractive. In addition, the authors have a concern about the possibility of mixed genotype (WT and het) in lines 445 – 447. Even though previous another project has reported no impact on results, is this speculation scientifically accepted? There is no guarantee whether heterozygous was used or as WT or not.

2. The representative images will help to understand the quantified data in Figures 3B – E (mitochondria), Figure 5 (Ribbon), and Figure 7F (data from Opa1 mutant). In Figure 3, the reviewer could not consider whether the data from 3 – 6 dpf can be combined without the images. In addition to the image, please consider the statistical method in Figures 5A – C, although the conclusion won't be affected. The authors analyzed the data by one-way ANOVA with Tukey's test, however, two-way ANOVA will be applicable in this case because central and peripheral cells are paired. Figure 7F and F' seem to be identical. Please confirm the images.

3. The overall data, the authors analyzed using the cells from 3, 5, 6 dpf. Some data was pooled from all developmental stages and others were separated. It should be consistent throughout the study. The authors explained that 3 dpf neuromasts did not yet have a concentric patterning (lines 144 – 145). If so, how do the authors distinguish central and peripheral in Figure 5H? If they can distinguish, the data of 3 dpf in Figure 5H can be split into peripheral and central to fit the same style with 5 and 6dpf data. While 5 and 6 dpf data are treated differently depending on the data. The data from 5 and 6 dpf are mostly analyzed separately but not in Figure 5H. Moreover, the regression curve to show the correlation between kinocilium length and mitochondria or ribbon was only 5 dpf. If the author wants to discuss the development, the data 3, 5, and 6 dpf need to be analyzed separately and all data should be in the manuscript as main or supplemental figures.

4. The description of Figures 8F and G in the text (lines 245 – 246) does not match to the figure. The text says two-way ANOVA but the figure legend says Kruskal-Wallis test. Adding WT data to Figure 8E (even if it is the same data as Figure 2) and performing Two-way ANOVA will help the reader to understand.

5. In the Discussion section, the authors are trying to interpret their morphological data by comparing previous functional analyses, however, the discussion includes many speculations about the mitochondrial function. In addition to TMRE and Ca measurement in Opa1 mutant, the author needs to address some of the functions using their models if they want to focus on the mitochondrial functions. For instance ROS measurement, ATP production, biogenesis markers analysis, or response to neomycin as described in the introduction section comparing each developmental stage (3, 5, 6 dpf) or among cell types (hair cell, central, peripheral…).

Esterberg et al., J Clin Invest. 2016 (https://www.jci.org/articles/view/84939)

Alassaf et al., eLife, 2019 (https://elifesciences.org/articles/47061#s2)

Hirose et al., Hearing Research, 2016 (https://www.sciencedirect.com/science/article/pii/S0378595516301599?via%3Dihub)

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "Activity regulates a cell type-specific mitochondrial phenotype in zebrafish lateral line hair cells" for further consideration by eLife. Your revised article has been evaluated by Didier Stainier (Senior Editor) and a Reviewing Editor.

The manuscript has been much improved. While the reviewers agreed that no additional experiments are needed, specific suggestions for clarification are outlined below.

Reviewer #1 (Recommendations for the authors):

I have completed my review of the revised manuscript by McQuate et al.

The authors have done an admirable job addressing the suggestions provided by all the reviewers. On its own, this work is a well-rounded, complete body of work. In addition, this work represents an immense resource for the field. I do not have any major concerns moving forward-the paper is suitable for publication.

Reviewer #2 (Recommendations for the authors):

The consistency of the analysis and data presentation was much improved in this revised manuscript. Yet a detailed interpretation of what the UMAP analysis reveals about mitochondrial properties in relation to hair cell maturation and in the context of the two mutants was not included in this revision. It is important for the authors to unpack for the readers what they think their UMAP analysis reveals and how it supports their hypothesis.

- The authors did not adequately interpret their UMAP analysis in Figure 5 nor explain what they think it reveals about mitochondrial properties in relation to hair cell maturation. Stating that older hair cells have "a significantly different mitochondrial phenotype than younger hair cells" is too vague. For example, there appear to be differences in the clustering of hair cell mitochondrial morphologies based on stereocilia length (Figure 5 A) and chronological age of the NM (Figure 5 B), yet there is an area of clear overlap in clustering on the left side where all three chronological ages overlap (Figure 5 B) in an area with shorter stereocilia length (Figure 5 A). This reviewer thinks that detail may provide the support that functional maturation at all three chronological ages plays a key role in establishing hair cell mitochondrial morphology, but if that's the case, it should be discussed by the authors in the results.

- Similarly, a more detailed interpretation is needed of what the UMAP analysis in Figure 9 reveals about the differences in cdh23-/- vs. cav1.3a-/- hair cell mitochondria.

- The terms "young" and "mature" (5 dpf) or "older" (3 dpf) to describe hair cells are used throughout the manuscript, yet the authors state there are no categorical values for maturity. Some measurable criteria (i.e. range of stereocilia lengths) must have been used for the authors to identify cells as "young" vs "older/mature", and these stereocilia length parameters should be defined in the methods.

- Better cohesion on the discussion of how calcium may be influencing the size of mitochondria is needed. In line 333 the authors identify calcium influx at the apical end as potentially driving the smaller size of mitochondria, yet don't address that the largest mitochondria are also found near synaptic ribbons with high levels of calcium influx until line 402. Speculation on how calcium may be influencing these two cellular compartments differently should be included in the discussion, but they should be in one section and discussed in relation to one another.

Reviewer #3 (Recommendations for the authors):

The revised manuscript has been improved a lot and is well aligned.

Data and figures are revised to make the readers understand easily.

The reviewer found one critical mislabel in Figures 6B and C.

WT and Opa1 data seem to be switched. So, the reviewer would like the author to confirm it.

Finally, even though the author analyzed many mitochondria, the author did not increase the number of opa1 mutant fish (used only 1 fish for analysis) but emphasized the data robustness. However, the reviewer is still wondering whether the data from n = 1 can indicate robustness, support the scientific meaning, and convince the reader.

The eLife transparent reporting form asks authors to state information about sample sizes (which should include details of the methods used to estimate them and the assumptions made).

https://elifesciences.org/articles/21070 eLife journal confirms and admits, it will be no big problem.

Essential revisions:

All three reviewers thought this high-resolution mitochondrial study in sensory hair cells is novel and interesting. Nevertheless, there are some valid concerns raised by the reviewers that should be improved such as:

1) A better link between mitochondrial disruption to hair cell function in opa1 mutants,

We have completed new experiments using a cytoplasmic calcium indicator in addition to the mitochondrial calcium indicator for waterjet experiments comparing WT and opa1 mutants. We find that there is no significant difference between WT and mutants in cytoplasmic calcium changes in response to stimulation, suggesting that mechanotransduction is not substantially altered. By contrast, mitochondrial calcium increases after stimulation is significantly reduced in mutants compared to WT.

2) Improve data analyses, and

We have made major changes to improve data analysis and better support our overall conclusions. (1) We removed data from partially sectioned neuromasts and added additional data from newly reconstructed complete neuromasts so that all analyses could be consistent throughout. (2) We analyzed additional neuromasts from more fish to ensure our data is representative. (3) We re-analyzed mitochondrial changes related to hair cell maturation by including new measurements of stereocilia length, previously shown by Kindt et al. 2012 to be a robust marker of hair cell age analogous to kinocilium length. We include correlative analysis of stereocilia length and kinocilium length supporting this previous study, and now consistently apply this metric across our current study. We therefore no longer use categorical variables such as stage or position as a proxy for age. (4) We re-organized figures to improve consistency throughout the paper so that it is easier to compare wildtype and mutant animals. While we believe that our analysis is much more robust, our overall conclusions are unchanged from the previous submission.

3) Better functional evidence to support the hypothesis based on morphological analysis.

We certainly recognize that there are many additional functional studies suggested by our current morphological analysis. We believe that the functional data we do include has been improved (see comment 1 above). However, our current study is by no means insubstantial and indeed is unprecedented in its detail. We believe it will serve as an important resource for additional studies.

We will be happy to look at a revision of this manuscript should you be able to address the reviewers' concerns.

Reviewer #1 (Recommendations for the authors):

This is a rich dataset and resource for the community that has been thoroughly and thoughtfully analyzed. The presentation of the data and figures is high quality and thoughtfully executed. Our comments seek to further strengthen the manuscript.

1. Please provide a table(s) to summarize the EM sample data, including the age, neuromast identity/identities, genotype, whether the samples are whole or partial neuromasts, which cells have been reconstructed (HCs, SCs), and a list of the figures that use data from that sample.

As noted above we now include data only from whole neuromasts, and summarize this data collection in Table 1. We have completely reorganized figures to make clear which datasets are used in which figures. In addition, we provide the names of each dataset and how they precisely relate to each figure in the supplementary material. We further note that these datasets have been made publicly available. We also include all source data.

2. Include information in your introduction that highlights the total number of samples reconstructed and the public availability of the data. This is an amazing resource for the community and is worth highlighting more.

We thank the Reviewer for appreciating the sheer magnitude of work that went into this study, and its usefulness for the scientific community. Overall, we reconstructed 5,908 mitochondria from 162 cells from 16 neuromasts from 10 fish. We added this summary in a Table, along with a statement of public availability in the introduction (lines 90-99).

3. For cdh23 and cav1.3 mutants, please provide a comparison of wt to mutant mitochondrial volume/total cell volume as in Figure S1B. It looks like the HC cell size may be different in some of the mutants, which may impact mitochondrial volume.

We now include cell volumes for cdh23 mutants (Figure 7C) and cav1.3 mutants (Figure 8C). While HCs are on average slightly larger in mutants than in WT, there is no statistical difference in the ratio of mitochondrial volume to cell volume in these different genotypes.

4. Sample sizes. How many different fish were analyzed for P-SCs? The number of NMs and fish are different in different figure legends. If n = 1 fish, there should be more than 1 NM and more than 1 fish analyzed for P-SC mitos. It would seem that the data already exists (as there are 2+ fish for both HCs and C-SCs), so it should be possible to at least increase these numbers to 2 NMs and 2 fish for all cell types. More broadly, multiple fish would ideally be examined for each mutant. We recognize that this is incredibly labor-intensive work, but recent EM datasets published in eLife have shown data for n>1 even in mutants (ex. Dow et al. 2018).

We now include data for both peripheral and central SCs from 3 neuromasts in 2 different fish and note this in the text (line 112).

We now include data for cdh23 and cav1.3a mutants from 4 different neuromasts in 2 different fish, and note this in the text-

cdh23: line 252.

cav1.3a: line 270.

It is an honor to have our work compared to the landmark Dow et al. 2018 paper. This study was primarily concerned with the HC body, and thus only contains reconstructions of whole HCs. Our manuscript differentiates from the previous study by individually reconstructing each mitochondrion within HCs (on average around 40 per HC), often in addition to the cell body. The manual reconstruction and individual quantification of each mitochondrion is where the bulk of the laborious nature of our study lies.

5. Can you show how the position of the max mito changes over developmental time for Figures 3 and 4?

We now show changes in the position of the max mito during development for 5-6 dpf WT (Figure 2H), 3 dpf WT (Figure 4H), cdh23 (Figure 7H), and cav1.3a (Figure 8H).

6. In a supplement it would be helpful to add a plot showing the sizes of all mitochondria for a mature, representative HC(s)? It would be nice to see an example(s) of the full distribution.

We’ve included this in Figure 2—figure supplement 2.

7. Opa1 mutations have a broad range of effects, as mitochondrial fusion is hypothesized to impact many aspects of mitochondrial function. This makes it hard to link functional changes in these mutants to changes in mito architecture. Please discuss this caveat.

We thank the reviewer for pointing out this important caveat when interrupting mitochondrial fusion, and have added it to the discussion (Lines 342-343).

8. Although evoked mitoGCaMP3 responses in opa-1 mutants are compelling, changes in evoked responses could also be due to defects in mechanotransduction. Another method to verify normal mechanotransduction would be helpful. For example, normal FM 1-43 labeling or unaltered evoked cyto- or mem-GCaMP6 responses.

We thank the reviewer for suggesting this important control. We have now repeated waterjet experiments measuring both cytoplasmic and mitochondrial calcium responses in wildtype and mutant animals (Figure 6). We find no significant difference in cytoplasmic calcium responses between mutant and WT animals, but see a marked reduction in mitochondrial calcium responses in mutants compared to WT.

Reviewer #2 (Recommendations for the authors):

Results and Figures:

– Labeling of larval age and hair cell age and position is inconsistent throughout the results and figures. For example, in Figure 2 A neither larval age nor hair cell position are specified.

We thank the Reviewer for pointing this out. As summarized above, we have completely redesigned this analysis of the relationships between HC maturity and mitochondrial and ribbon phenotypes, and we believe it is now more consistent and clear.

– Line 139-140: "Younger hair cells on periphery and mature hair cells in center" need a citation.

We have removed the analysis of position and maturity.

– Development of mitochondria with maturation (Figure 3): the authors show images of representative neuromasts from younger (3 dpf) and more mature (5-6 dpf) neuromasts in Figure 3A, yet analyzed pooled mitochondrial morphologies from all three ages using the position of the hair cell in the neuromast as a proxy for maturity. Given that 3 dpf neuromasts are not yet functionally mature and do not show concentric patterning comparable to older neuromasts, it is unclear why the authors used this approach. The analysis shown in Figure 3B-E would be more convincing if central and peripheral hair cell data were shown for just 5-6 dpf, with 3 dpf shown separately.

We thank the Reviewer for noting the confusing presentation of the data in the previous submission. The 5-6dpf data are consistently pooled throughout the manuscript, with the 3 dpf data having its own separate figure. The only places all three ages are pooled are in the UMAP analysis, where we are specifically analyzing the HC mito developmental trajectory.

– Figure 4A: How were young, intermediate, and mature hair cells defined?

We no longer assign categorical values for maturity.

– Figure 5D-F: Is this data pooled from all 3 ages?

This analysis has changed.

– UMAP data/ Figure 6 and Figure 10: A more detailed description of UMAP data analysis and what it specifically reveals about mitochondrial properties in relation to hair cell maturation and/or function is needed.

We have added more description of the analysis to the manuscript text (lines 201-204) and additional details of how the test was conducted to the Methods (lines 509-518).

– opa1 mutants/ Figure 7: The authors show reduced evoked mitochondrial calcium uptake in opa1 mutants. Did they examine mitoGCaMP levels at rest (i.e. compare mitoGCaMP fluorescence in WT vs mutant without stimulation)?

We thank the Reviewer for pointing out baseline calcium mitochondrial levels could be different between WT and mutant. We now include a comparison baseline fluorescence in Figure 6 —figure supplement 1, and show there was no significant difference between WT and mutant.

Additionally, there is a possibility that the hair cells in opa1 mutants have impaired mechanotransduction. Did the authors examine FM1-43 uptake in mutants vs. WT?

As noted in the response to Reviewer 1, we have now analyzed both cytoplasmic and mitochondrial calcium responses in mutants, and show that mutants have no significant impairment of mechanotransduction.

– cdh23 mutants/ Figure 8: To support the conclusion that the largest mitochondria do not localize to the presynaptic ribbons, the authors should show the localization of the ribbons in the representative reconstructions in Figure 8A.

We thank the Reviewer for pointing out how to make this conclusion stronger. Images for all genotypes now include ribbons.

– cav1.3a mutants/ Figure 9: This analysis overall would be far more impactful if it corresponded with the analysis performed on cdh23 mutants. The authors should repeat the analysis of the data in the same way it was performed for cdh23 mutants, then expand on the cav1.3a mutant analysis with an additional figure.

We have completely revised the analysis to be more comprehensive. Moreover we have made figure layouts of wildtype and mutant analysis more consistent.

– UMAP data/ Figure 10: In Figure 10A, does the WT condition contain cells from all ages (3 and 5/6 dpf)? If so, that additional variable may be confounding, as the two mutants examined only represent hair cells from 5 dpf fish.

Removal of the 3 dpf data does not impact the results of the UMAP (now included as Figure 9 —figure supplement 2). We’ve included the 3 dpf data in the main figure so that it sheds light on the developmental trajectory.

Discussion:

– The discussion on the possible mechanisms for distinct mitochondrial morphology at the apical and basolateral ends of hair cells needs some revising. It's unclear to this reviewer how the data shown supports that calcium signaling shapes mitochondrial morphology through fission. The authors propose that calcium influx through MET may itself contribute to smaller mitochondria, yet large amounts of calcium enter presynaptic Cav1.3a channels when active hair cells are stimulated, and this is where the mitochondrial volume is the largest. Conversely, while the authors suggest that large mitochondrial networks near the synapses may be specifically necessary to drive PMCA1 pumps, it should be noted that PMCA2 pumps play a critical role in the regulation of calcium at hair cell stereocilia. Overall, it seems that other key differences, such as the high metabolic demand of synaptic transmission itself, may play a role in the localization of large mitochondria to synapses.

We thank the Reviewer for pointing out this interesting paradox! We have softened these speculations to indicate calcium pumps may be only one of several aspects of synaptic transmission needing metabolic support, and further emphasized the metabolic demand of synaptic transmission possibly driving max mito formation (lines 345-349).

Materials and methods:

– More details are needed for waterjet and calcium imaging. What concentration of MESAB was used? What was the composition of the extracellular solution? Which neuromast was acquired?

Added (Lines 480-491). The composition of the external solution (EM) is listed in line 439-440.

Reviewer #3 (Recommendations for the authors):

1. Some data from the too small sample size. In particular, P-SC in Figure 1 and 2,

We now include additional supporting cell data from more cells across several fish.

and Opa1 mutant in Figure 6A, and B are from only 1 fish.

We now emphasize that the opa1 phenotype is robust and visible through fluorescence, as seen in the supplemental figure. We believe the EM reconstruction is sufficient to make our point.

Although the authors pointed out the laborious nature of the reconstructions and suggested that their data reflect the greater population in the Discussion section, the reviewer believes that their reconstruction method and results are the central topics of this study. Therefore, the description of too much difficulty and speculation will make their promising data and techniques unattractive.

We have re-edited the discussion to soften this point.

In addition, the authors have a concern about the possibility of mixed genotype (WT and het) in lines 445 – 447. Even though previous another project has reported no impact on results, is this speculation scientifically accepted? There is no guarantee whether heterozygous was used or as WT or not.

We have removed these data and replaced it with reconstruction of a new neuromast that is WT. For the Reviewer’s interest, we note that there was no change in the conclusion.

2. The representative images will help to understand the quantified data in Figures 3B – E (mitochondria), Figure 5 (Ribbon), and Figure 7F (data from Opa1 mutant). In Figure 3, the reviewer could not consider whether the data from 3 – 6 dpf can be combined without the images.

We thank the Reviewer for pointing out how to strengthen the presentation of these data. We have completely re-done figures and believe that the images now clearly represent the quantified data.

In addition to the image, please consider the statistical method in Figures 5A – C, although the conclusion won't be affected. The authors analyzed the data by one-way ANOVA with Tukey's test, however, two-way ANOVA will be applicable in this case because central and peripheral cells are paired.

We have changed this analysis and no longer use categorical labels.

Figure 7F and F' seem to be identical. Please confirm the images.

We note that these images were NOT identical. However, we have now replaced them with other images to make this point clearer to the casual reader.

3. The overall data, the authors analyzed using the cells from 3, 5, 6 dpf. Some data was pooled from all developmental stages and others were separated. It should be consistent throughout the study. The authors explained that 3 dpf neuromasts did not yet have a concentric patterning (lines 144 – 145). If so, how do the authors distinguish central and peripheral in Figure 5H? If they can distinguish, the data of 3 dpf in Figure 5H can be split into peripheral and central to fit the same style with 5 and 6dpf data. While 5 and 6 dpf data are treated differently depending on the data. The data from 5 and 6 dpf are mostly analyzed separately but not in Figure 5H. Moreover, the regression curve to show the correlation between kinocilium length and mitochondria or ribbon was only 5 dpf. If the author wants to discuss the development, the data 3, 5, and 6 dpf need to be analyzed separately and all data should be in the manuscript as main or supplemental figures.

We recognize that the previous presentation was confusing and therefore made extensive changes to the analysis. Foremost is the use of stereocilia length to measure hair cell maturity, which vacated the requirement for categorical analysis. We removed data from incomplete neuromasts and added new data to expand the analysis. We now directly compare data from different age fish as well to strengthen our conclusions. All data is included in the manuscript, and in addition we include all source data. We note that none of our conclusions have changed from the previous submission.

4. The description of Figures 8F and G in the text (lines 245 – 246) does not match to the figure. The text says two-way ANOVA but the figure legend says Kruskal-Wallis test. Adding WT data to Figure 8E (even if it is the same data as Figure 2) and performing Two-way ANOVA will help the reader to understand.

We thank the Reviewer for pointing out that this was confusing. Given the non-normality of these data, we opted to change this analysis to only comparing differences in max-mito percentage in the basal-most HC quadrant using a non-parametric test, which we deemed more appropriate than a parametric two-way ANOVA. We included the new comparison in all relevant figures.

5. In the Discussion section, the authors are trying to interpret their morphological data by comparing previous functional analyses, however, the discussion includes many speculations about the mitochondrial function.

We believe that interpreting our data in light of previous published studies is not excessively speculative. We believe we have been both balanced and accurate. Indeed, comparing reported results to other studies is the fundamental underlying function of a Discussion section of any paper.

In addition to TMRE and Ca measurement in Opa1 mutant, the author needs to address some of the functions using their models if they want to focus on the mitochondrial functions. For instance ROS measurement, ATP production, biogenesis markers analysis, or response to neomycin as described in the introduction section comparing each developmental stage (3, 5, 6 dpf) or among cell types (hair cell, central, peripheral…).

Esterberg et al., J Clin Invest. 2016 (https://www.jci.org/articles/view/84939)

Alassaf et al., eLife, 2019 (https://elifesciences.org/articles/47061#s2)

Hirose et al., Hearing Research, 2016 (https://www.sciencedirect.com/science/article/pii/S0378595516301599?via%3Dihub)

While we agree with the Reviewer that these additional studies would be interesting, they are beyond the scope of the current paper. We believe that the manuscript as it stands includes substantial data and analysis that we certainly hope will spark future study.

[Editors’ note: further revisions were suggested prior to acceptance, as described below.]

Reviewer #2 (Recommendations for the authors):

The consistency of the analysis and data presentation was much improved in this revised manuscript. Yet a detailed interpretation of what the UMAP analysis reveals about mitochondrial properties in relation to hair cell maturation and in the context of the two mutants was not included in this revision. It is important for the authors to unpack for the readers what they think their UMAP analysis reveals and how it supports their hypothesis.

- The authors did not adequately interpret their UMAP analysis in Figure 5 nor explain what they think it reveals about mitochondrial properties in relation to hair cell maturation. Stating that older hair cells have "a significantly different mitochondrial phenotype than younger hair cells" is too vague.

- Similarly, a more detailed interpretation is needed of what the UMAP analysis in Figure 9 reveals about the differences in cdh23-/- vs. cav1.3a-/- hair cell mitochondria.

We thank the Reviewer for the comment, and have added additional text to clarify why we performed PCA analysis and accompanying UMAP dimensionality reduction. We realize that separating the PCA analyses into two separate figures was potentially confusing to readers, so instead we have combined Figures 5 and 9 into one figure.

Our overall goal was to test whether we saw a distinction between wildtype and mutant HCs with respect to mitochondrial properties when we examined all these properties in aggregate. However these measurements likely have co-variance, so we used PCA, as principal components are by definition uncorrelated. We then represented the multidimensional PCA space as a projection in two dimensions using UMAP. Finally we performed spatial autocorrelation analysis to test whether stereocilia length was differentially distributed across the UMAP projection and then used neighbor analysis to test whether genotypes differentially clustered.

We have combined figures 5 and 9 into one figure (new Figure 8). Additional text describing the rationale and analysis ins found in lines 275-286.

For example, there appear to be differences in the clustering of hair cell mitochondrial morphologies based on stereocilia length (Figure 5 A) and chronological age of the NM (Figure 5 B), yet there is an area of clear overlap in clustering on the left side where all three chronological ages overlap (Figure 5 B) in an area with shorter stereocilia length (Figure 5 A). This reviewer thinks that detail may provide the support that functional maturation at all three chronological ages plays a key role in establishing hair cell mitochondrial morphology, but if that's the case, it should be discussed by the authors in the results.

We provided evidence that there are hair cells with varying degrees of maturity at different developmental stages when comparing 3dpf to 5-6dpf hair cells in Figure 4. As suggested, we have added additional text here to again highlight this point.

- The terms "young" and "mature" (5 dpf) or "older" (3 dpf) to describe hair cells are used throughout the manuscript, yet the authors state there are no categorical values for maturity. Some measurable criteria (i.e. range of stereocilia lengths) must have been used for the authors to identify cells as "young" vs "older/mature", and these stereocilia length parameters should be defined in the methods.

This information can be found in Figure 2- Supplement Figure 1, Figure 4-Supplement Figure 1, and Figure 8-Supplement Figure 1. Stereocilia length was used as a continuous measure of HC age throughout the manuscript, and the labels directly reflect the linear regression shown in the aforementioned Figures, which is why the stereocilia length for HCs in the Main Figures is included alongside the labels “young” or “mature.” We have added a line to the methods to clarify this point.

- Better cohesion on the discussion of how calcium may be influencing the size of mitochondria is needed. In line 333 the authors identify calcium influx at the apical end as potentially driving the smaller size of mitochondria, yet don't address that the largest mitochondria are also found near synaptic ribbons with high levels of calcium influx until line 402. Speculation on how calcium may be influencing these two cellular compartments differently should be included in the discussion, but they should be in one section and discussed in relation to one another.

The two points are addressed in sequential paragraphs, and the conundrum is pointed out in line 408:

“This suggests a paradox where calcium might promote both mitochondrial fission apically and fusion basally. However, synaptic transmission requires large energy expenditures in addition to calcium regulation, and these demands may instead drive mitochondrial fusion into basal networks. Understanding the paradoxical effects of HC activity influencing both the formation of smaller apical mitochondria and larger basal networks will require additional study.”

Reviewer #3 (Recommendations for the authors):

The revised manuscript has been improved a lot and is well aligned.

Data and figures are revised to make the readers understand easily.

The reviewer found one critical mislabel in Figures 6B and C.

WT and Opa1 data seem to be switched. So, the reviewer would like the author to confirm it.

We thank the reviewer for pointing out this mistake. We’ve corrected it.

Finally, even though the author analyzed many mitochondria, the author did not increase the number of opa1 mutant fish (used only 1 fish for analysis) but emphasized the data robustness. However, the reviewer is still wondering whether the data from n = 1 can indicate robustness, support the scientific meaning, and convince the reader.

The eLife transparent reporting form asks authors to state information about sample sizes (which should include details of the methods used to estimate them and the assumptions made).

https://elifesciences.org/articles/21070 eLife journal confirms and admits, it will be no big problem.

We clarify (lines 201- 208) that mutations in opa1 cause mitochondrial fragmentation from yeast to humans. We state that the zebrafish phenotype is readily observable and completely penetrant by fluorescent imaging. The n = 1 fish included 5 hair cells, and is in regards to the SBF reconstruction, which is only included in this figure to make the point definitive. Physiology experiments used multiple neuromasts from multiple fish.

Author details

1. Andrea McQuate

   1. Department of Biological Structure, University of Washington, Seattle, United States
   2. Department of Otolaryngology-HNS, University of Washington, Seattle, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Funding acquisition, Investigation, Methodology, Writing – original draft, Project administration

   For correspondence

   amcquate@uw.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-2052-2004

2. Sharmon Knecht

   Department of Biological Structure, University of Washington, Seattle, United States

   Contribution

   Resources

   Competing interests

   No competing interests declared

3. David W Raible

   1. Department of Biological Structure, University of Washington, Seattle, United States
   2. Department of Otolaryngology-HNS, University of Washington, Seattle, United States

   Contribution

   Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Investigation, Methodology, Project administration, Writing - review and editing

   For correspondence

   draible@uw.edu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-5342-5841

• 1,335

  Page views

• 195

  Downloads

• 1

  Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.