1. Cell Biology
  2. Neuroscience

Pregnancy-associated plasma protein-aa regulates endoplasmic reticulum–mitochondria associations

  1. Mroj Alassaf
  2. Mary C Halloran  Is a corresponding author
  1. Department of Integrative Biology, University of Wisconsin, United States
  2. Department of Neuroscience, University of Wisconsin, United States
  3. Neuroscience Training Program, University of Wisconsin, United States
https://doi.org/10.7554/eLife.59687
Share this article
Cite this article
  1. Mroj Alassaf
  2. Mary C Halloran
(2021)
Pregnancy-associated plasma protein-aa regulates endoplasmic reticulum–mitochondria associations
eLife 10:e59687.
https://doi.org/10.7554/eLife.59687
  2. Download BibTeX
  3. Download .RIS

Abstract

Endoplasmic reticulum (ER) and mitochondria form close physical associations to facilitate calcium transfer, thereby regulating mitochondrial function. Neurons with high metabolic demands, such as sensory hair cells, are especially dependent on precisely regulated ER–mitochondria associations. We previously showed that the secreted metalloprotease pregnancy-associated plasma protein-aa (Pappaa) regulates mitochondrial function in zebrafish lateral line hair cells (Alassaf et al., 2019). Here, we show that pappaa mutant hair cells exhibit excessive and abnormally close ER–mitochondria associations, suggesting increased ER–mitochondria calcium transfer. pappaa mutant hair cells are more vulnerable to pharmacological induction of ER–calcium transfer. Additionally, pappaa mutant hair cells display ER stress and dysfunctional downstream processes of the ER–mitochondria axis including altered mitochondrial morphology and reduced autophagy. We further show that Pappaa influences ER–calcium transfer and autophagy via its ability to stimulate insulin-like growth factor-1 bioavailability. Together our results identify Pappaa as a novel regulator of the ER–mitochondria axis.

Introduction

Neuronal survival is critically dependent on proper mitochondrial function, which is regulated in part by close associations between mitochondria and the endoplasmic reticulum (ER). Despite the diverse pathologies of neurodegenerative diseases, increasing evidence suggests that a disrupted ER–mitochondria connection is a common underlying feature (Area-Gomez et al., 2012; Bernard-Marissal et al., 2015; Calì et al., 2013; Hedskog et al., 2013). Indeed, many of the cellular processes associated with neurodegeneration such as mitochondrial dysfunction, mitochondrial fragmentation, disrupted autophagy, and oxidative stress are regulated by the ER–mitochondria axis (Csordás et al., 2018; Friedman et al., 2011; Rowland and Voeltz, 2012).

Precise regulation of mitochondrial calcium levels is essential for mitochondrial metabolism and cell survival (Cárdenas et al., 2010). The association of mitochondria with ER facilitates efficient uptake of calcium by the mitochondria (Krols et al., 2016; Rizzuto et al., 1998). Calcium released from the ER forms highly concentrated calcium microdomains that are necessary to override the low affinity of the mitochondrial calcium uniporter (MCU), the primary mitochondrial calcium uptake channel (Rizzuto et al., 2012). However, overly close inter-organelle distances or excessive ER–mitochondria contact points can result in increased calcium transfer, which in turn can overstimulate mitochondrial bioenergetics leading to the production of pathological levels of the energy byproduct, reactive oxygen species (ROS). Accumulation of ROS can then lead to oxidative stress, fragmentation of the mitochondrial network, and cell death (Adam-Vizi and Starkov, 2010; Brookes et al., 2004; Deniaud et al., 2008; Esterberg et al., 2014; Iqbal and Hood, 2014; Ježek et al., 2018). Thus, tight regulation of the distance and frequency of contacts between the ER and mitochondria is critical for proper mitochondrial calcium load. Previous studies investigating regulation of ER–mitochondria associations have focused on identifying proteins involved in tethering the ER to the mitochondria. However, the upstream molecular pathways that act to prevent excessive and pathologically tight ER–mitochondria associations, and the mechanisms by which these molecular pathways are regulated remain unknown.

To investigate these mechanisms, we are using zebrafish lateral line hair cells, which share molecular, functional, and morphological similarities with mammalian inner ear hair cells. Hair cells of the inner ear and lateral line system are specialized sensory neurons that transduce acoustic information and relay it to the central nervous system (McPherson, 2018). Sensory hair cells are an excellent model to investigate mechanisms regulating ER–mitochondria associations. Their high metabolic requirements make them particularly vulnerable to disruptions in mitochondrial calcium load (Gonzalez-Gonzalez, 2017). Indeed, excess mobilization of calcium from the ER to the mitochondria was recently shown to underlie aminoglycoside-induced hair cell death in the zebrafish lateral line (Esterberg et al., 2014). These findings highlight the importance of the ER–mitochondria connection in regulating hair cell survival. Given that hair cell death causes hearing loss (Eggermont, 2017), identifying molecular factors that regulate the ER–mitochondria connection may provide insight into potential therapeutic targets to combat ototoxin-induced hearing loss.

We showed previously that pregnancy-associated plasma protein-aa (Pappaa) regulates mitochondrial function in zebrafish lateral line hair cells (Alassaf et al., 2019). Pappaa is a locally secreted metalloprotease that regulates Insulin-like growth factor 1 (IGF1) signaling. IGF1 is sequestered extracellularly by IGF binding proteins (IGFBPs) that prevent it from binding to its receptor. Pappaa cleaves inhibitory IGFBs, thereby stimulating local IGF1 availability and signaling (Hwa et al., 1999). We found that loss of Pappaa, and the consequential reduction in IGF1 signaling, resulted in mitochondrial calcium overload, ROS buildup, and increased susceptibility to ROS-induced cell death (Alassaf et al., 2019). Moreover, other studies have shown that declining IGF1 levels correlate with conditions such as aging, neurodegenerative disorders, and obesity, in which the ER–mitochondria connection is known to be altered (Arruda et al., 2014; Hedskog et al., 2013; Lee et al., 2018; Liu and Zhu, 2017). However, a causative link between IGF1 signaling and ER–mitochondria associations has not been shown.

In this study, we reveal that Pappaa, an extracellular regulator of IGF1 signaling, is critical for regulation of the ER–mitochondria connection. We use electron microscopy (EM) to show that ER–mitochondria associations are closer and greater in number in pappaa mutant hair cells. Hair cells in pappaa mutants are more sensitive to pharmacological induction of ER-mediated calcium release and show changes in mitochondrial morphology and stunted autophagic response. Loss of Pappaa also results in ER stress and activation of the unfolded protein response. Together, our results suggest that Pappaa exerts its effect on hair cell survival by serving as a key regulator of the ER–mitochondria axis. Given the widespread roles for IGF1 receptors and the suppressive effects of IGFBPs, exogenous IGF1 treatment may not be an effective therapeutic approach for neurodegenerative diseases. A factor such as Pappaa that can locally activate IGF1 signaling may provide a more promising therapeutic target to prevent hearing loss.

Results and discussion

Pappaa influences ER–mitochondria associations

Zebrafish lateral line hair cells lie on the surface of the skin and are arranged into structures called neuromasts (Figure 1A,B). Each neuromast consists of a cluster of hair cells surrounded by the glia-like support cells. Our previous work showed that hair cells in pappaa mutants (hereafter referred to as pappaap170) are more susceptible to ROS-induced death and their mitochondria have increased calcium load (Alassaf et al., 2019). Therefore, we hypothesized that pappaap170 mitochondria may have an increased frequency of ER-mitochondria associations. To test this, we used EM to visualize ER-mitochondria associations. We collected 80 nm thick sections along the apical-basal axis of anterior lateral line neuromasts of 5 days post fertilization (dpf) wild-type and pappaap170 larvae (Figure 1A,B). Hair cells were identified based on their central location and darker cytoplasm as previously described (Behra et al., 2009; Owens et al., 2007; Suli et al., 2016). We quantified the number of ER tubules within 100 nm of mitochondria, the maximum distance for effective mitochondrial calcium uptake (Csordás et al., 2018). The ER was identified by its ribosome-rich membrane and elongated shape (blue profiles in Figure 1C). Because the ER at post-synaptic sites adjacent to efferent inputs has a distinct role in buffering high levels of post-synaptic calcium influx, and thus may impact the ER–mitochondria axis differently (Moglie et al., 2018), we did not include ER within 100 nm of the post-synaptic sites in our analysis. To identify the post-synaptic ER, we used the efferent terminal as a landmark, which presents as an electron light body filled with vesicles (Figure 1—figure supplement 1A–A’). We also did not include ER in EM sections with presynaptic terminals, identified by synaptic ribbons (Figure 1—figure supplement 1B–B’). We found that pappaap170 hair cells had on average 34% more ER tubules in close association with each mitochondrion (Figure 1C,D). Furthermore, we quantified the percentage of mitochondria in close association with a given number of ER tubules. We found that 61% of mitochondria in pappaap170 hair cells had one or more ER tubules in close association compared to 40% of wild-type hair cell mitochondria, and 25% of mutant mitochondria had two or more associated ER tubules compared with only 9% in wild type (Figure 1C,E). These results suggest that Pappaa regulates the frequency of ER–mitochondria associations.

Figure 1 with 1 supplement see all
Download asset Open asset
Pappaa regulates ER-mitochondria associations.

(A–A’) Schematic of zebrafish lateral line hair cells. (A) The dotted lines represent EM plane of section. (A’) Schematic of a dorsal view (left) and lateral view (right) of a single neuromast. (B) Representative EM section of lateral line neuromast taken along the apical-basal axis of lateral line hair cells (blue) in 5 dpf larva. Scale bar = 4 μm. (C) Representative EM images of ER-mitochondria associations in wild-type and pappaa hair cells. ER is pseudo colored in blue. Scale bar = 200 nm. (D) Mean number of ER tubules within 100 nm of mitochondria. ****p<0.0001 t-test, Mann–Whitney correction. N = 459 mitochondria (wild type) and 447 mitochondria (pappaap170) collected from six larvae/genotype. Error bars = SEM. (E) Percentages of mitochondria associated with 0, 1, 2, 3, or 4 ER tubules. *p<0.05 chi-square test. N = 459 mitochondria (wild type) and 447 mitochondria (pappaap170) collected from six larvae/genotype. (F) KDEL immunolabeling in 5 dpf wild-type and pappaap170 brn3c:mGFP-labeled hair cells. (G) Mean percentage of area covered by KDEL immunolabeling per neuromast. Unpaired t-test with Welch correction revealed no significant difference between groups p=0.84. N = 15–17 larvae/genotype (shown at base of bars), 1–3 neuromasts/larva. Total number of neuromasts included in the analysis = 35 (wild type) and 31 (pappaap170) neuromasts from two experiments. (H) Mean length of ER tubules. t-test with Mann–Whitney correction found no significant difference. N = 119 ER tubules (wild type) and 123 ER tubules (pappaap170) collected from six larvae/genotype. (I) Mean distance between the ER and mitochondria that are within 100 nm of each other. *p<0.05 t-test, Mann–Whitney correction. N = 131 ER-mitochondria associations (wild type) and 234 ER-mitochondria associations (pappaap170) collected from six larvae/genotype. Error bars=SEM.

Figure 1—source data 1

Mean number of ER tubules/mitochondrion.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig1-data1-v2.xlsx
Figure 1—source data 2

Percentages of mitochondria that are associated with 0, 1, 2, 3, or 4 ER tubules.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig1-data2-v2.xlsx
Figure 1—source data 3

Mean percentage of area covered by KDEL immunolabeling.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig1-data3-v2.xlsx
Figure 1—source data 4

Mean ER tubule length.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig1-data4-v2.xlsx
Figure 1—source data 5

Mean ER-mitochondria distance.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig1-data5-v2.xlsx

We next asked whether the increase in the number of ER–mitochondria associations in pappaap170 hair cells was due to an overabundance of ER. We used an antibody shown to specifically label KDEL (Blanco-Sánchez et al., 2018; Blanco-Sánchez et al., 2014), an ER C-terminal peptide retention signal (Munro and Pelham, 1987), to visualize the ER in the transgenic line Tg(brn3c:mGFP), in which the lateral line hair cells are labeled with membrane-targeted GFP (Figure 1F). We measured the percentage of area occupied by KDEL immunofluorescence per neuromast using an approach previously described (Blanco-Sánchez et al., 2014). We found that the mean KDEL % area per neuromast was not different between wild-type and pappaap170 larvae, suggesting there is no ER expansion in pappaap170 hair cells (Figure 1F–G). As another means to assess ER abundance, we measured the length of ER tubules in our EM images. We found that there is no difference in the length of individual ER tubules (Figure 1H). Notably, the total number of ER tubules in this analysis did not significantly differ between wild type (119 tubules) and pappaap170 (123 tubules) hair cells (Figure 1H). These results, in addition to our previous findings showing no difference in mitochondrial mass between wild-type and pappaap170 hair cells (Alassaf et al., 2019), indicate that the increase in the number of ER-mitochondrion associations we see in pappaap170 hair cells (Figure 1D,E) is not simply due to increased ER or mitochondrial mass, but instead reflects a specific effect on promoting close associations.

The distance between ER and mitochondria also influences the efficiency of mitochondrial calcium intake and must be tightly regulated (Rizzuto et al., 2012). Thus, we asked whether pappaap170 hair cells also had tighter ER–mitochondria associations. We measured the distances between all ER and mitochondria within 100 nm of each other. We found that the ER-mitochondria distance was on average 5 nm shorter in pappaap170 hair cells (Figure 1C,I). Taken together, our data suggest that Pappaa is important for preventing excessive and overly tight ER–mitochondria associations. Interestingly, another study used a synthetic linker to shorten the inter-organelle distance and showed a surge in mitochondria calcium load (Csordás et al., 2006), suggesting that the excessively tight connections we see in pappaap170 hair cells could be causing their increased mitochondrial calcium load. Given the novelty of Pappaa’s role in regulating ER-mitochondria associations, it will be interesting to identify the downstream molecular targets of Pappaa within the ER–mitochondria tethering complex.

Pappaa loss sensitizes hair cells to increased ER–mitochondria calcium transfer

Given the increased ER–mitochondria associations in pappaap170 hair cells, we hypothesized that they experience excessive ER–mitochondria calcium transfer. If so, we would expect that pappaap170 hair cells would show increased vulnerability to pharmacological stimulation of ER-mitochondria calcium transfer. To address this question, we used thapsigargin, a sarco/endoplasmic reticulum Ca²+-ATPase (SERCA) inhibitor, which prevents ER calcium influx and thereby allows for more available calcium in the cytosol to be taken up by the mitochondria (Esterberg et al., 2014; Figure 2A). A previous study used a transgenic zebrafish line with a mitochondria-targeted calcium indicator (mitoGCaMP3) to show that thapsigargin treatment caused increased mitochondrial calcium uptake in lateral line hair cells (Esterberg et al., 2014). We treated wild-type and pappaap170 larvae with thapsigargin and used the Tg(brn3c:mGFP) transgene to assess hair cell survival. We used a series of thapsigargin doses, either chronic exposure to low concentrations or acute exposure to high concentrations. We found that treatment for 24 hr with low doses of thapsigargin caused 11–21% more hair cell death in pappaap170 larvae compared to wild-type larvae (Figure 2C). Similarly, we found that 1 hr exposure to relatively high doses of thapsigargin at 5 dpf resulted in 13–22% more hair cell death in pappaap170 larvae (Figure 2B,D). These findings suggest that pappaap170 hair cells are more sensitive to fluctuations in calcium and further support the idea that pappaap170 hair cells have a disrupted ER–mitochondria connection that results in greater mitochondrial calcium uptake. However, it is difficult to discern whether the elevated levels of calcium in pappaap170 mitochondria are solely due to the altered ER-mitochondria associations. It is possible that pappaap170 hair cells experience increased ER calcium efflux and/or mitochondrial calcium uptake. Protein kinase B (Akt), which is a downstream effector of multiple signaling pathways including IGF1, has been shown to inhibit IP3Rs and reduce ER calcium release in HeLa cells (Marchi et al., 2008). If Akt has a similar role in hair cells, it is possible that Pappaa-IGF1 signaling may normally act to stimulate Akt-induced suppression of IP3Rs, preventing excessive ER calcium release. Further investigation into the activity of ER and mitochondria associated calcium channels in pappaap170 hair cells is needed to answer this question.

Figure 2 with 1 supplement see all
Download asset Open asset
pappaap170 hair cells are more sensitive to disruption in ER–mitochondria calcium signaling.

(A) Thapsigargin increases calcium concentration at the ER-mitochondria junction by blocking the SERCA pump and inhibiting calcium uptake by the ER. (B) Representative images of brn3c:mGFP-labeled hair cells from vehicle or 10 µM thapsigargin-treated larvae. (C) Mean percentage of surviving hair cells following 24 hr treatments with thapsigargin starting at 4 dpf. To calculate hair cell survival percentage, hair cell number post-drug treatment was normalized to mean hair cell number in vehicle treated larvae of the same genotype. **p<0.01, ****p<0.0001, two-way ANOVA, Holm–Sidak post-test. N = 8–13 larvae per group (shown at base of bars), three neuromasts/larva were analyzed. Total number of neuromasts included in the analysis = 24 (wild type; vehicle-treated), 24 (pappaap170; vehicle-treated), 30 (wild type; 0.3 μM thapsigargin), 33 (pappaap170; 0.3 μM thapsigargin), 30 (wild type; 0.7 μM thapsigargin), 24 (pappaap170; 0.7 μM thapsigargin), 39 (wild type; 1 μM thapsigargin), 36 (pappaap170; 1 μM thapsigargin). (D) Mean percentage of surviving hair cells following 1 hr treatment with thapsigargin at 5 dpf. *p<0.05, **p<0.01, two-way ANOVA, Holm–Sidak post-test. N = 8–13 larvae per group (shown at base of bars), three neuromasts/larva from two experiments were analyzed. Total number of neuromasts included in the analysis = 60 (wild type; vehicle-treated), 60 (pappaap170; vehicle-treated), 36 (wild type; 10 μM thapsigargin), 33 (pappaap170; 10 μM thapsigargin), 39 (wild type; 20 μM thapsigargin), 36 (pappaap170; 20 μM thapsigargin). (E) Mean percentage of surviving hair cells following induction of dnIGF1R expression. To calculate hair cell survival percentage, hair cell number after 1 hr treatments with thapsigargin was normalized to mean hair cell number in non-heat-shocked, vehicle-treated larvae of the same genotype. ***p<0.001 two-way ANOVA, Holm–Sidak post-test. N = 8–10 larvae per group (shown at base of bars), three neuromasts per larva. Total number of neuromasts included in the analysis = 30 (wild type; non-heat-shocked, vehicle-treated), 27 (dnIGF1Ra; non-heat-shocked, vehicle-treated), 30 (wild type; heat-shocked, vehicle-treated), 30 (dnIGF1Ra; heat-shocked, vehicle-treated), 24 (wild type; non-heat-shocked, 10 μM thapsigargin), 30 (dnIGF1Ra; non-heat-shocked, 10 μM thapsigargin), 27 (wild type; heat-shocked, 10 μM thapsigargin), 27 (dnIGF1Ra; heat-shocked, 10 μM thapsigargin), 30 (wild type; non heat-shocked, 20 μM thapsigargin), 30 (dnIGF1Ra; non-heat-shocked, 20 μM thapsigargin), 27 (wild type; heat-shocked, 20 μM thapsigargin), 27 (dnIGF1Ra; heat-shocked, 20 μM thapsigargin). (F) Mean percentage of surviving hair cells following co-treatment with NBI-31772 and 20 µM thapsigargin. To calculate hair cell survival percentage, hair cell counts after treatment were normalized to hair cell number in vehicle treated larvae of the same genotype. *p<0.05, **p<0.01. Two-way ANOVA, Holm–Sidak post-test. N = 9–13 larvae per group (shown at base of bars), three neuromasts per larva. Total number of neuromasts included in the analysis = 24 (wild type; vehicle-treated), 24 (pappaap170; vehicle-treated), 39 (wild type; 20 μM thapsigargin), 36 (pappaap170; 20 μM thapsigargin), 27 (wild type; 20 μM thapsigargin+ 100 μM NBΙ−31772), 39 (pappaap170; 20 μM thapsigargin + 100 μM NBΙ−31772), 33 (wild type; 20 μM thapsigargin + 120 μM NBΙ−31772), and 33 (pappaap170; 20 μM Thapsigargin + 120 μM NBΙ−31772).

Figure 2—source data 1

Hair cell survival following 24 hr treatment with thapsigargin in wild-type and pappaap170 larvae.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig2-data1-v2.xlsx
Figure 2—source data 2

Hair cell survival following 1 hr treatment with thapsigargin in wild-type and pappaap170 larvae.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig2-data2-v2.xlsx
Figure 2—source data 3

Hair cell survival following induction of dnIGF1R expression.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig2-data3-v2.xlsx
Figure 2—source data 4

Hair cell survival following co-treatment with thapsigargin and NBI-31772 in wild-type and pappaap170 larvae.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig2-data4-v2.xlsx

We sought to define the molecular pathway by which Pappaa regulates ER-mitochondria calcium transfer. Pappaa is a secreted metalloprotease and a known regulator of IGF1 signaling. Through its proteolytic activity, Pappaa cleaves the inhibitory IGFBPs, freeing IGF to bind to its receptor and thereby increasing IGF bioavailability and signaling (Boldt and Conover, 2007). To determine whether Pappaa regulates ER–mitochondria calcium transfer by stimulating IGF1 signaling, we took two approaches. First, we reasoned that if IGF1 signaling plays a role in regulating ER–mitochondria calcium transfer, then attenuating IGF1 signaling would result in increased sensitivity of wild-type hair cells to thapsigargin. Second, we reasoned that pharmacological stimulation of IGF1 bioavailability in pappaap170 mutants would improve hair cell survival in response to thapsigargin. To attenuate IGF1 signaling, we used a transgenic line with an inducible heat shock promoter that drives the expression of a dominant negative IGF1 receptor [Tg(hsp70:dnIGF1Ra-GFP)] (Kamei et al., 2011). Given that we previously showed IGF1 signaling acts post-developmentally to support hair cell survival (Alassaf et al., 2019), we induced dnIGF1R expression on day four and exposed the larvae to thapsigargin for 1 hr. Notably, heat shock alone in wild-type or Tg(hsp70:dnIGF1Ra-GFP) larvae did not alter hair cell survival (Figure 2E). However, expression of dnIGF1R did exacerbate hair cell loss in response to thapsigargin. Interestingly, 10 μM thapsigargin did not kill wild-type hair cells; however, it did cause hair cell death in dnIGF1Ra-GFP expressing larvae (Figure 2E), similar to pappaap170 larvae. This suggests that IGF1R signaling attenuates ER–mitochondria calcium transfer. To stimulate IGF1 bioavailability in pappaap170 mutants, we treated larvae with NBI-31772, an IGFFBP inhibitor, for 24 hr and then exposed the larvae to a 1 hr treatment with thapsigargin. We found that treatment with NBI-31772 significantly improved hair cell survival in pappaap170 larvae (Figure 2F). These data support the idea that Pappaa regulates ER-mitochondria calcium transfer via its ability to increase the bioavailability of IGF1.

Pappaa is not expressed by the hair cells themselves, rather it is expressed by the surrounding support cells of the neuromast (Alassaf et al., 2019). Because Pappaa is a secreted protein, it could potentially affect IGF signaling in the hair cells, support cells, or both. To ask which cells show altered IGF1 signaling in pappaap170 mutants, we used an antibody against phosphorylated IGF1R (pIGF1R) that was previously validated using an IGF1R inhibitor (Chablais and Jazwinska, 2010). IGF1R undergoes autophosphorylation when it binds to IGF1, thereby activating its signaling cascade (Feldman et al., 1997). We quantified the fluorescent intensity of pIGF1R in wild-type and pappaap170 neuromasts. We found that loss of Pappaa caused a substantial reduction in pIGF1R fluorescence in the hair cells (Figure 2—figure supplement 1A–B), suggesting that Pappaa secreted by support cells can impact IGF1R signaling in the neighboring hair cells. Interestingly, the support cells in pappaap170 mutants also showed reduced pIGF1R fluorescence (Figure 2—figure supplement 1A,C), suggesting that Pappaa can have an autocrine effect on support cells. It is possible that hair cells are also affected indirectly by reduced IGF signaling in support cells. In support of this idea, activation of phosphoinositide 3-kinase (PI3K), which is downstream of IGF1R, in the support cells of mammalian cochlea resulted in enhanced survival of hair cells against cisplatin, a chemotherapy agent, suggesting that IGF1 signaling in support cells could influence hair cells. However, the mechanism by which PI3K signaling in support cells ultimately affects cochlear hair cells is not fully understood (Jadali et al., 2017).

Pappaa regulates mitochondrial morphology

Mitochondria undergo constant fission and fusion as a form of quality control. Damaged mitochondria are excised from the mitochondrial network through the process of fission and cleared by autophagy, while the remaining reusable pool of mitochondria fuse back together (Ni et al., 2015). Thus, a balance between mitochondrial fission and fusion must be achieved to maintain cellular homeostasis (Liesa et al., 2009). Under pathological conditions, such as oxidative stress in which there is an excess of damaged mitochondria, the balance between fission and fusion tips toward fission. Excessive fission can lead to fragmentation of the mitochondrial network and ultimately to cell death. Recently, it was shown that mitochondrial fission occurs at the site of ER-mitochondria contacts. ER tubules were found to wrap around mitochondria and mark the site for fission (Friedman et al., 2011). Given that pappaap170 hair cells experience oxidative stress, have dysfunctional mitochondria (Alassaf et al., 2019), and exhibit increased ER-mitochondria associations, we hypothesized that pappaap170 hair cells experience increased fission. Mitochondrial morphology can be predictive of fission and fusion events. Small, round, and clustered mitochondria are often indicative of fission events, whereas large and branched mitochondria are indicative of fusion events. To test our hypothesis, we used EM and measured several parameters of mitochondrial morphology that are typically used as a readout of mitochondrial fragmentation (Wiemerslage and Lee, 2016). These parameters include the area, perimeter, aspect ratio (major axis/minor axis; a measure of elongation), circularity (4π × area/perimeter2), and network interconnectivity (area/perimeter) (Wiemerslage and Lee, 2016). We found that pappaap170 mitochondria had a smaller area, perimeter, and aspect ratio than wild-type mitochondria. Moreover, they were more circular and showed reduced network interconnectivity, suggesting excessive mitochondrial fission and possible network fragmentation (Figure 3A–F). To further investigate mitochondrial integrity in whole cells, we labeled hair cells with the vital mitochondrial dye Mitotracker. Using confocal microscopy, we acquired z-stacks of neuromast hair cells. Mitochondria in pappaap170 hair cells appeared less connected and more rounded compared to wild type (Figure 3G, Videos 1 and 2). Analysis of mitochondrial circularity using ImageJ (Figure 3H) revealed results similar to the EM data, supporting the idea that pappaap170 mitochondria are fragmented. Taken together, our data suggest that Pappaa regulates mitochondrial fission. Our findings are consistent with a previous study showing that treatment with IGF1 maintained mitochondrial integrity in denervated muscle cells by promoting mitochondrial fusion and preventing mitochondrial fission (Ding et al., 2017).

Figure 3
Download asset Open asset
Pappaa loss causes mitochondrial fragmentation.

(A) Representative EM images of mitochondria in lateral line hair cells in 5 dpf wild-type and pappaap170 larvae. (B–F) Mean mitochondrial (B) area, (C) perimeter, (D) circularity, (E) aspect ratio, and (F) interconnectivity in 5 dpf wild-type and pappaap170 lateral line hair cells. ****p<0.0001 t-test, Mann–Whitney correction. N = 272 mitochondria (wild type) and 262 mitochondria (pappaap170) collected from six larvae/genotype. (G) Representative images of 5 dpf wild-type and pappaap170 lateral line hair cells loaded with the vital mitochondrial dye, Mitotracker. Images are maximum intensity projection through neuromast in xy view, with cross sections of yz plane shown at right and xz plane shown at bottom. (H) Mean mitochondrial circularity measured from Z-stack max intensity projections of wild-type and pappaap170 lateral line hair cells. ****p<0.0001 t-test, Welch correction. N = 8 larvae per group (shown at base of bars), one neuromast/ larva. Error bars=SEM. See Videos 1 and 2.

Figure 3—source data 1

Mitochondrial area in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data1-v2.xlsx
Figure 3—source data 2

Mitochondrial perimeter in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data2-v2.xlsx
Figure 3—source data 3

Mitochondrial circularity in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data3-v2.xlsx
Figure 3—source data 4

Mitochondrial aspect ratio in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data4-v2.xlsx
Figure 3—source data 5

Mitochondrial interconnectivity in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data5-v2.xlsx
Figure 3—source data 6

Mitochondrial circularity in wild-type and pappaap170 lateral line hair cells measured by mitotracker.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig3-data6-v2.xlsx
Video 1
Download asset
3D rendering of mitotracker labeling in wild-type lateral line hair cells.

This video was constructed from a confocal z-stack of wild-type lateral line hair cells loaded with mitotracker.

Video 2
Download asset
3D rendering of mitotracker labeling in pappaap170 lateral line hair cells.

This video was constructed from a z-stack of pappaap170 lateral line hair cells loaded with mitotracker.

Pappaa regulates neomycin-induced autophagy

We next asked if other processes known to be regulated by the ER–mitochondria axis are disrupted in pappaap170 hair cells. Multiple studies suggest that autophagy, a bulk degradation process, is strongly influenced by the ER–mitochondria axis. Pharmacological and genetic manipulations that tighten or increase ER–mitochondria associations result in reduced autophagosome formation (Gomez-Suaga et al., 2017a; Gomez-Suaga et al., 2017b; Sano et al., 2012). Therefore, we asked whether pappaap170 hair cells have an attenuated autophagic response. Interestingly, neomycin, to which pappaap170 hair cells show hypersensitivity (Alassaf et al., 2019), was shown to be sequestered into autophagosomes upon entry into hair cells (Hailey et al., 2017; He et al., 2017). These authors also showed that blocking autophagy exacerbates neomycin-induced damage presumably due to the heightened exposure of vital intracellular components to neomycin. Thus, neomycin uptake into autophagosomes can be used as an indicator of the robustness of the autophagic response. To test whether pappaap170 hair cells show a dampened autophagic response to neomycin treatment, we used a fluorescently tagged neomycin (Neomycin-Texas Red). Upon cell entry through the mechanotransduction (MET) channels (Alharazneh et al., 2011; Hailey et al., 2017; Kroese et al., 1989), neomycin-Texas Red appears as a diffused pool in the cytosol. Within minutes, neomycin gets captured by autophagosomes and high fluorescent intensity puncta begin to appear (Hailey et al., 2017; Figure 4 and Video 3). We counted the number of puncta within visually accessible neuromast hair cells in wild-type and pappaap170 larvae. We found that wild-type hair cells had 300% more neomycin-Texas red puncta. (Figure 4B,C and Video 4), indicating that neomycin-induced autophagy was attenuated in pappaap170 hair cells. This reduction in autophagy may allow neomycin to accumulate in the cytosol, causing more damage to intracellular compartments. To test the possibility that the increased neomycin-induced hair cell death in pappaap170 larvae (Alassaf et al., 2019) is the result of enhanced entry of neomycin through the MET, we measured the change in fluorescent intensity of neomycin-Texas Red (ΔF/F0) in individual hair cells at 2.5, 4.5, and 6.5 min post neomycin-Texas Red exposure. We found that pappaap170 hair cells displayed similar neomycin-Texas Red ΔF/F0 to wild-type hair cells at every time point (Figure 4D). To determine whether pappaap170 hair cells experience overall increased neomycin-Texas Red loading, we looked at the maximum ΔF/F0 and found no difference between pappaap170 and wild-type hair cells (Figure 4E). This is consistent with our previous finding showing that MET channel-mediated entry of the fluorescent dye FM1-43 was not altered in pappaap170 hair cells (Alassaf et al., 2019).

Figure 4
Download asset Open asset
Pappaa regulates neomycin-induced autophagy.

(A) Schematic showing cell entry and autophagy of neomycin-Texas Red. (B) Representative time lapse images of brn3C:mGFP-labeled neuromast hair cells (green) at 5 dpf following exposure to 10 μM neomycin-Texas Red (white). (C) Mean number of neomycin-Texas Red puncta/hair cell in wild-type and pappaap170 larvae at 5 dpf. **p<0.01 t-test, Mann–Whitney correction. N = 19 hair cells (wild type) and 22 hair cells (pappaap170) collected from four larvae/genotype. (D) Mean neomycin-Texas Red ΔF/F0 at 2.5, 4.5, and 6.5 min post-exposure. Multiple t-test with Holm–Sidak correction found no significant difference. N = 22 hair cells (wild type) and 22 hair cells (pappaap170) collected from four larvae/genotype. (E) Maximum change in neomycin-Texas Red fluorescent intensity across treatment time. Unpaired t-test with Mann–Whitney correction found no significant difference. N = 22 hair cells (wild type) and 22 hair cells (pappaap170) collected from four larvae/genotype. See Videos 1 and 2. (F) Representative time lapse images of vehicle or 120 μM NBI-31772 treated brn3C:mGFP-labeled neuromast hair cells (green) of pappaap170 larvae at 5 dpf following exposure to 10 μM neomycin-Texas Red (white). (G) Mean number of neomycin-Texas Red puncta/hair cell in vehicle or 120 μM NBI-31772 treated pappaap170 larvae at 5 dpf. ****p<0.0001 t-test, Mann–Whitney correction. N = 30 hair cells (Vehicle) and 24 hair cells (120 μM NBI-31772) collected from four larvae/ group. Error bars=SEM.

Figure 4—source data 1

Mean number of neomycin-Texas Red puncta in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig4-data1-v2.xlsx
Figure 4—source data 2

Mean change in neomycin-Texas Red fluorescence over time in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig4-data2-v2.xlsx
Figure 4—source data 3

Maximum change in neomycin-Texas Red fluorescence in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig4-data3-v2.xlsx
Figure 4—source data 4

Mean number of neomycin-Texas Red puncta in NBI-31772-treated pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig4-data4-v2.xlsx
Video 3
Download asset
Wild-type lateral line hair cells following exposure to neomycin-Texas red.

This video was constructed from time lapse images of wild-type brn3c:mGFP lateral line hair cells (left) following exposure to 10 μM neomycin-Texas red (right).

Video 4
Download asset
pappaap170 lateral line hair cells following exposure to neomycin-Texas red.

This video was constructed from time lapse images of pappaap170 brn3c:mGFP lateral line hair cells (left) following exposure to 10 μM neomycin-Texas red (right).

We next asked whether Pappaa regulates autophagy by increasing the availability of IGF1. To answer this, we treated pappaap170 larvae with 120 μM NBI-31772 for 24 hr starting at 4 dpf before exposure to neomycin-Texas red at 5 dpf. Compared to vehicle treated pappaap170 controls, which show an average of 1 neomycin-Texas Red punctate/hair cell, treatment with NBI-31772 resulted in a threefold increase in the number of neomycin-Texas Red puncta (Figure 4F–G), which is similar to wild-type levels (Figure 4C). Together, these findings suggest that Pappaa regulates neomycin-induced autophagy by increasing the availability of IGF1.

Pappaa loss causes ER stress

ER stress can be a potential cause of mitochondria dysfunction. In addition to acting as the largest calcium reserve in the cell, the ER is the primary hub for protein processing. Newly synthesized proteins enter the ER for chaperone-assisted folding (Gregersen et al., 2006). Improper or insufficient folding results in the accumulation of unfolded proteins, which causes ER stress and increased calcium efflux (Houck et al., 2012). This efflux results in an increase in highly concentrated calcium ‘hot spots’ that act as arresting signals for mitochondria, tethering the mitochondria to the ER and thereby increasing the frequency and tightness of ER–mitochondria associations (Bravo et al., 2012; Bravo et al., 2011). To ask whether pappaap170 hair cells experience ER stress, we FACsorted hair cells from wild-type and pappaap170 Tg(brn3c:mGFP) larvae and evaluated gene expression of the unfolded protein response (UPR) by RT-qPCR. The UPR consists of three ER transmembrane receptors that act as sensors for unfolded proteins, IRE1, ATF4, and PERK (Figure 5A). Normally, Bip, an ER-resident chaperone, is bound to the UPR receptors. When unfolded proteins begin to accumulate, Bip dissociates from the UPR receptors to assist in protein folding. The uncoupling of Bip activates the UPR signaling cascade signifying ER stress. During the early phase of ER stress, the UPR promotes cell survival by improving the folding capacity of the ER through the upregulation of pro-survival factors including bip, atf4, and the splicing of xbp1. However, a switch from an adaptive to a pro-apoptotic UPR occurs during the late phase of ER stress, in which chop, a pro-apoptotic transcription factor, is upregulated (Fu et al., 2015; Oslowski and Urano, 2011). Our analysis revealed that components of the ‘adaptive’ UPR pathway (bip, atf4, and spliced-xbp1) were upregulated in pappaap170 hair cells (Figure 5B). Notably, the ‘pro-apoptotic’ factor, chop, was not differentially expressed in pappaap170 hair cells. These results suggest that pappaap170 hair cells experience early ER stress but not the later, pro-apoptotic response. To further investigate whether pappaap170 hair cells are in the process of apoptosis, we performed terminal deoxynucleotidyl transferase dUTP nick-end labeling (TUNEL) and used phalloidin as a counterstain to identify hair cells. As a positive control, we treated wild-type larvae with neomycin, a known apoptosis-inducing drug, for 30 min prior to fixation, as previously described (Goodman and Zallocchi, 2017; Figure 5C). In all neuromasts analyzed (head neuromasts in 10 larvae/group), we observed multiple TUNEL-positive hair cells in the positive controls. In contrast, in both untreated wild-type and pappaap170 larvae, we did not observe any TUNEL-positive neuromasts, which is consistent with the lack of spontaneous hair cell death in pappaap170 larvae (Alassaf et al., 2019). Together with our finding that pappaap170 hair cells do not exhibit ER expansion (Figure 1C,F), an event associated with advanced ER stress as a result of extreme accumulation of misfolded proteins in the ER (Chavez-Valdez et al., 2016; Dorner et al., 1989; Welihinda et al., 1999), these data support the idea that the ER in pappaap170 hair cells is in the early stages of stress. Interestingly, another study using a zebrafish model of Usher syndrome, a genetic disorder that can cause hearing loss (Blanco-Sánchez et al., 2014), found that while Usher mutant larvae did not show spontaneous hair cell death, the hair cells were TUNEL positive and showed ER expansion, suggesting that Usher mutants experience more advanced stages of ER stress than pappaap170 mutants.

Figure 5
Download asset Open asset
Pappaa loss causes ER stress.

(A) Schematic of the UPR pathway. The accumulation of unfolded proteins activates the UPR receptors, IRE1, ATF6, and PERK, signifying ER stress. In the early adaptive phase of ER stress, the UPR promotes cell survival through the upregulation of pro-survival factors including bip, atf4, and spliced xbp1. A switch from an adaptive to a pro-apoptotic UPR occurs during the late phase of ER stress in which Chop, a pro-apoptotic transcription factor, is upregulated. (B) Mean fold change in UPR mRNA levels in wild-type and pappaap170 hair cells at 5 dpf. N = 2–3 technical replicates/gene. *p<0.05, ***p<0.001, ****p<0.0001, two-way ANOVA, Holm–Sidak post-test. Error bars=SEM. (C) Representative images of TUNEL staining (magenta) in wild-type and pappaap170 lateral line hair cells. Stereocilia are counterstained with phalloidin (white). A 30 min treatment with 100 μM neomycin was used as positive control (top). (D) Mean percentage of surviving hair cells following a 24 hr treatment with tunicamycin starting from 4 dpf. To calculate hair cell survival percentage, hair cell number post-treatment was normalized to mean hair cell number in vehicle-treated larvae of the same genotype. *p<0.05, **p<0.01, two-way ANOVA, Holm–Sidak post-test. N = 8–10 larvae per group (shown at base of bars), three neuromasts/larva from two experiments were analyzed. Total number of neuromasts included in the analysis = 51 (wild type; vehicle treated), 57 (pappaap170; vehicle treated), 27 (wild type; 2 μM Tunicamycin), 27 (pappaap170; 2 μM Tunicamycin), 27 (wild type; 3 μM Tunicamycin), 27 (pappaap170; 3 μM Tunicamycin). Error bars=SEM.

Figure 5—source data 1

Mean fold change in UPR transcript levels in wild-type and pappaap170 lateral line hair cells.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig5-data1-v2.xlsx
Figure 5—source data 2

Hair cell survival following treatment with Tunicamycin in wild-type and pappaap170 larvae.

https://cdn.elifesciences.org/articles/59687/elife-59687-fig5-data2-v2.xlsx

To ask whether pappaap170 hair cells are more vulnerable to pharmacological induction of protein unfolding given their already active UPR, we treated 5 dpf wild-type and pappaap170 Tg(brn3c:mGFP) larvae with tunicamycin, an inhibitor of glycoprotein biosynthesis (Oslowski and Urano, 2011), for 24 hr, and analyzed hair cell survival. pappaap170 larvae showed 20% less hair cell survival compared to wild type, suggesting that they are closer to the cytotoxic threshold of ER stress (Figure 5D). Together, these findings suggest that Pappaa is essential for ER homeostasis.

Conclusion

The ER and mitochondria are physically and functionally linked together to facilitate a range of cellular processes essential for neuron survival. Disruptions to this connection can cause damage to vital cellular components and is found to occur in many neurodegenerative diseases. To identify potential targets for therapeutic interventions, we must define and characterize the genetic and molecular mechanisms regulating the ER–mitochondria axis. Here, we describe a novel role for Pappaa, a local stimulator of IGF1 signaling, in regulating the ER–mitochondria axis and its essential downstream processes. We previously showed that mitochondria in pappaa mutant hair cells are dysfunctional, resulting in oxidative stress (Alassaf et al., 2019). Here we provide additional mechanistic insight into Pappaa’s role in hair cell survival. We show that Pappaa is essential for the function of ER and its relationship to mitochondria in hair cells. Hair cell function and activity are influenced by the surrounding support cells as well as the efferent and afferent neurons that innervate hair cells (Babola et al., 2020; Faucherre et al., 2009; Pichler and Lagnado, 2020). A caveat to our pharmacological experiments is that drug treatments will affect all of these cell types, and some effects on hair cells may be indirect. Nonetheless, we show that the ER stress, excessive and abnormally tight ER–mitochondria associations, defects in mitochondria morphology and reduced autophagic activity in pappaap170 mutants occur in the hair cells. A challenge to determining the initial site of dysfunction within pappaap170 hair cells is the functional coupling of the ER and mitochondria. ER-mediated calcium release acts as a key regulator of mitochondrial bioenergetics, and thereby ROS production (Görlach et al., 2015). In turn, mitochondria-generated ROS can stimulate the activity of ER-associated calcium channels causing further calcium release (Brookes et al., 2004; Chaudhari et al., 2014; Görlach et al., 2015). Given that many ER-resident molecular chaperones are calcium dependent, ER stress can be both a cause and a consequence of mitochondrial dysfunction (Zhu and Lee, 2015). While it is difficult to definitively determine which organelle is affected first when the ER–mitochondria axis is compromised, several of our findings support an ER-driven dysfunction in pappaap170 mutants. Bravo et al., 2011 showed that during early ER stress, increased ER–mitochondria coupling occurs to enhance mitochondrial bioenergetics. pappaap170 hair cells display all the telltale signs of early ER stress, including activation of the adaptive UPR branch (Figure 5B), heightened mitochondrial bioenergetics evident by increased calcium load, ROS, and hyperpolarization (Alassaf et al., 2019). Furthermore, Esterberg et al., 2014 characterized the timeline of subcellular events following ER calcium release in hair cells. They showed that calcium released from the ER is quickly taken up by the mitochondria and initially causes mitochondria hyperpolarization suggesting increased bioenergetics. However, once a certain threshold is crossed, depolarization of the mitochondria occurs, which is followed by permeabilization of the outer membrane and eventually cell death, hence this step is often referred to as ‘the point of no return’. pappaap170 mitochondria are hyperpolarized (Alassaf et al., 2019), which places them a step after ER–calcium release and before depolarization of the outer mitochondrial membrane and hair cell death. Indeed, pappaap170 hair cells do not spontaneously degenerate (Figure 5C). Although it is difficult to determine with absolute certainty, we speculate that ER stress may precede mitochondrial dysfunction in pappaap170 hair cells. Precise temporal dissection is needed to identify Pappaa’s primary subcellular target, whether the mitochondria, ER, or both.

IGF1 emerges as a promising therapeutic candidate due to its known role in influencing the function of the ER and the mitochondria. Studies using exogenous IGF1 treatment showed it can alleviate pharmacologically induced ER stress in cancer cells (Novosyadlyy et al., 2008) or attenuate mitochondria-generated ROS in cirrhosis or aging disease models (Castilla-Cortazar et al., 1997; García-Fernández et al., 2008; Novosyadlyy et al., 2008; Sádaba et al., 2016), supporting the possibility that IGF1 signaling may play a role in regulating the ER-mitochondria axis. However, exogenous IGF1 treatment presents its own set of challenges for developing an IGF-based therapy. IGF1 signaling has a broad cellular impact and ubiquitous overexpression could affect many systems. Additionally, the suppressive effects of IGFBPs on IGF1 may render this approach inefficient. Indeed, patients with neurodegenerative disorders have not shown significant improvement following systemic IGF1 administration. These disappointing outcomes are thought to be due to the suppressive effects of IGFBPs on IGF1 bioavailability (Raoul and Aebischer, 2004; Sakowski et al., 2009). This was evident in patients with amyotrophic lateral sclerosis, a progressive motor neuron disease (Bowling et al., 1993; Shaw et al., 1995), in which total IGF1 levels appeared to be normal, whereas ‘free’ IGF1 levels were reduced (Wilczak et al., 2003). As an upstream regulator of IGF1 signaling that is spatially restricted, Pappaa may provide a more viable therapeutic alternative.

Materials and methods

Key resources table
Reagent type
(species) or resource		DesignationSource or referenceIdentifiersAdditional information
Gene (Danio rerio)pappaap170PMID:25754827RRID:ZFIN_ZDB-GENO-190322-4Single-nucleotide nonsense mutation C>T at position 964 in Exon 3
Strain, strain background (Danio rerio)Tg(brn3c:GFP)PMID:15930106RRID:ZFIN_ZDB-ALT-050728-2
Strain, strain background (Danio rerio)TLFZebrafish International Resource Center (ZIRC)RRID:ZFIN_ZDB-GENO-990623-2
Antibodyanti-KDEL (mouse monoclonal)CalbiochemAB_2120901:500
Antibodyanti-GFP (rabbit polyclonal)ThermoFisher ScientificRRID:AB_2215691:500
AntibodyAlexa 488, secondary (rabbit polyclonal)ThermoFisher ScientificRRID:AB_25762171:500
Othermitotracker greenFMThermoFisher ScientificCatalog number: M7514 PubChem CID:70691021100 nM for 5 min
Other0.25% trypsin-EDTASigma-AldrichCatalog number: T3924
OtherTRIzolInvitrogenCatalog number: 15596026
OtherTexas RedThermoFisher ScientificCatalog number: T6134
OtherSso fast Eva Green SupermixBio-RadCatalog number: 1725200
Chemical compound, drugNeomycinSigma-AldrichCatalog number: N1142
Chemical compound, drugNBI-31772Fisher ScientificCatalog number: 519210
Chemical compound, drugThapsigarginTocrisCatalog number: 1138
Chemical compound, drugTunicamysinCaymen ChemicalCatalog number: 11445
Commercial assay or kitSuperScript II Reverse TranscriptaseInvitrogenCatalog number: 18064014
Software, algorithmFluoview (FV10-ASW 4.2)OlympusRRID:SCR_014215
Software, algorithmImarisBitplaneRRID:SCR_007370
Software, algorithmImageJPMID:22743772RRID:SCR_003070
Software, algorithmGraphPad PRISMgraphpadRRID:SCR_002798
Software, algorithmJMP Pro 15.0SAS Institute IncRRID:SCR_014242

Animals

To generate pappaa+/+ and pappaap170 larvae, adult pappaap170/+ zebrafish (on a mixed Tubingen long-fin [TLF], WIK background) were crossed into Tg(brn3c:mGFP)s356t and then incrossed. Embryonic and larval zebrafish were raised in E3 media (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2, 0.33 mM MgSO4, pH adjusted to 6.8–6.9 with NaHCO3) at 29°C on a 14 hr/10 hr light/dark cycle through 5 dpf (Gyda et al., 2012; Kimmel et al., 1995). pappaap170 larvae were obtained from crosses of pappaap170/+ adults. Mutant larvae were identified either by lack of swim bladder inflation (Wolman et al., 2015) or by genotyping after the experiment. For genotyping, PCR was performed as previously described (Wolman et al., 2015) using forward primer: AGACAGGGATGTGGAGTACG, and reverse primer: GTTGCAGACGACAGTACAGC. PCR conditions were as follows: 3 min at 94°C, followed by 40 cycles of 94°C for 30 s, 57°C for 1 min, and 70°C for 1 min. The PCR product was digested with MseI (New England Biolabs R0525S) resulting in a 245 bp mutant restriction fragment that is distinct from the 271 bp wild-type restriction fragment. The restriction digest reaction was run on a 3% agarose gel.

Pharmacology

Request a detailed protocol

All pharmacological experiments were performed on Tg(brn3c:mGFP) larvae. Compounds were added to the larval E3 media at 4–5 dpf. Thapsigargin (Tocris No. 1138; dissolved in DMSO) was added at 10–20 µM for 1 hr at 5 dpf or at 0.3–1 µM for 24 hr at 4 dpf. Tunicamycin (Caymen Chemical 11445; dissolved in DMSO) was added at 2–3 µM for 24 hr, beginning at 4 dpf. To stimulate IGF1 bioavailability, larvae were pre-treated with 100–120 µM NBI-31772 (Fisher Scientific 519210; dissolved in DMSO) for 24 hr starting from day 4, then exposed to 20 µM thapsigargin for 1 hr. Optimal drug concentrations and exposure times were determined from pilot experiments with a wide range of doses. The highest dose that yielded hair cell death but was not lethal to the larvae was chosen as the upper limit for the dose ranges used in experiments. Following each treatment period, larvae were washed three times with E3 before fixation with 4% paraformaldehyde (diluted to 4% w/v in phosphate buffered saline (PBS) from 16% w/v in 0.1 M phosphate buffer, pH 7.4). Vehicle-treated controls were treated with 0.1% DMSO in E3.

Induction of a dominant negative IGF1R

Request a detailed protocol

To attenuate IGF1R signaling, we used Tg(hsp70:dnIGF1Ra-GFP) larvae that express dnIGF1Ra-GFP under the control of zebrafish hsp70 promoter (Kamei et al., 2011). Expression of dnIGF1Ra-GFP was induced by a 1 hr heat shock at 37°C at 4.5 dpf followed by another heat shock 12 hr later at 5 dpf. To control for any possible heat shock effects, non-transgenic controls were subjected to the same treatment. Thapsigargin was added 3 hr after the second heat shock at 5 dpf.

Immunohistochemistry

Request a detailed protocol

To visualize hair cells, 5 dpf larvae were fixed in 4% paraformaldehyde for 1 hr at room temperature and then rinsed three times with PBS. Larvae were blocked for 1 hr at room temperature in incubation buffer (IB: 0.2% bovine serum albumin, 2% normal goat serum, 0.8% Triton-X, 1% DMSO, in PBS, pH 7.4). Primary antibodies included anti-GFP to visualize brn3c:mGFP hair cells (1:500, rabbit polyclonal; ThermoFisher Scientific, RRID:AB_221569), phosphorylated IGF1R anti-IGF1 receptor phospho Y1161, 1:100, rabbit IgG; Abcam; (Chablais and Jazwinska, 2010), and anti-KDEL to visualize the ER (1:500, mouse monoclonal; Calbiochem, RRID: AB_212090). Larvae were incubated in primary antibodies in IB overnight at 4°C. Larvae were then rinsed five times for 10 min with IB. Following that, larvae were incubated in fluorescently conjugated secondary antibodies in IB for 4 hr at room temperature and rinsed five times with IB. Secondary antibodies included AlexaFluor488-conjugated antibody (goat anti-rabbit polyclonal, 1:500; ThermoFisher Scientific, RRID:AB_2576217) and AlexaFluor594-conjugated secondary antibody (goat anti-rabbit polyclonal, 1:500; ThermoFisher Scientific RRID:AB_142057). Larvae were mounted in 70% glycerol in PBS. Images were acquired with an Olympus Fluoview confocal laser scanning microscope (FV1000) using Fluoview software (FV10-ASW 4.2).

Quantification of the percentage area occupied by KDEL immunofluorescence was done using ImageJ. Maximum-intensity projections of z-stacks that spanned the entire depth of the neuromast were generated. A binary mask for the KDEL channel was generated using automated Otsu thresholding values, and the area fraction was measured. pIGF1R fluorescence was measured using ImageJ (Schneider et al., 2012) by drawing a region of interest around brn3c:GFP-labaled hair cells or 10 randomly selected SOX2-labeled support cells of the neuromast from Z-stack summation projections that included the full depth of the neuromast. Background fluorescence intensity was measured by drawing a region of interest away from the neuromast in the same Z-stack summation projection. The corrected total cell fluorescence (CTCF) was used to subtract background fluorescence. The CTCF formula was as follows: Integrated Density − (Area of selected cells × Mean fluorescence of background) (McCloy et al., 2014).

Hair cell survival

Request a detailed protocol

Hair cell survival experiments were performed in Tg(brn3c:mGFP) and Tg(hsp70:dnIGF1Ra-GFP) larvae as previously described (Alassaf et al., 2019). Hair cells from three stereotypically positioned neuromasts (IO3, M2, and OP1) (Raible and Kruse, 2000) were counted from z-stacks and averaged for each larva. Hair cell survival was calculated as a percentage with the following formula: [(mean number of hair cells after treatment)/(mean number of hair cells in vehicle treated group)] × 100.

RT-qPCR

Request a detailed protocol

Fluorescent sorting of brn3c:mGFP hair cells, RNA extraction, cDNA synthesis, and RT-qPCR were performed as previously described (Alassaf et al., 2019). For each genotype, 200 five dpf Tg(brn3c:GFP) larvae were incubated for 15 min in Ringer’s solution (116 mM NaCl, 2.9 mM KCl, 1.8 mM CaCl2, 5 mM HEPES, pH 7.2) (Guille, 1999) prior to dissection. pappaap170 larvae were identified by lack of swim bladder inflation (Wolman et al., 2015). Tails were dissected and transferred into 1.5 mL tubes containing Ringer’s solution on ice. For cell dissociation, dissected tails were incubated in 1.3 mL of 0.25% trypsin–EDTA (Sigma-Aldrich) for 20 min and triturated gently by P1000 pipette tip every 5 min. To stop cell digestion, 200 µL of stop solution containing 30% fetal bovine serum and 6 mM CaCl2 in PBS solution was added to the samples (Steiner et al., 2014). The samples were centrifuged at 400 g for 5 min at 4°C, and the supernatant was removed. Ca2+-free Ringer’s solution (116 mM NaCl, 2.9 mM KCl, 5 mM HEPES, pH 7.2) was added to rinse the cell pellet, and the samples were centrifuged again. Finally, the cell pellet was resuspended in Ca2+-free Ringer’s solution and kept on ice until ready to FAC sort. Cells were filtered through a 40 µm cell strainer. A two-gate sorting strategy was used. Cell vitality was assessed by DAPI, followed by forward scatter and GFP gate to isolate GFP+ cells. Live and GFP+ cells were collected into RNAse-free tubes containing 500 µL of TRIzol reagent (Invitrogen) for total RNA extraction. cDNA was synthesized using SuperScript II Reverse Transcriptase (Invitrogen 18064014). Real-time quantitative PCR (RT-qPCR) was performed using Sso fast Eva Green Supermix (Bio-Rad 1725200) in a StepOnePlus Real-Time PCR System (Applied Biosystems) based on manufacture recommendation. Reactions were run in three to four technical replicates containing cDNA from 50 ng of total RNA/reaction. The primer sequences for the UPR genes are as follows: for bip, forward: ATCAGATCTGGCCAAAATGC and reverse: CCACGTATGACGGAGTGATG; atf4, forward: TTAGCGATTGCTCCGATAGC and reverse: GCTGCGGTTTTATTCTGCTC; for chop, forward: ATATACTGGGCTCCGACACG and reverse: GATGAGGTGTTCTCCGTGGT (Howarth et al., 2014); for xbp1-spliced, forward: TGTTGCGAGACAAGACGA and reverse: CCTGCACCTGCTGCGGACT (Vacaru et al., 2014); for b-actin (endogenous control), forward TACAGCTTCACCACCACAGC and reverse: AAGGAAGGCTGGAAGAGAGC (Wang et al., 2005 ). Cycling conditions were as follows: 1 min at 95°C, then 40 cycles of 15 s at 95°C, followed by 1 min at 60°C (Jin et al., 2010). Analysis of relative gene expression was done using the 2−ΔΔCt method (Livak and Schmittgen, 2001).

MitoTracker

Request a detailed protocol

To assess mitochondrial morphology, 5 dpf larvae were incubated in 100 nM mitotracker green FM (ThermoFisher Scientific M7514; dissolved in anhydrous DMSO) for 5 min. Following the incubation period, larvae were washed three times in E3, anesthetized in 0.02% tricaine (Sigma-Aldrich) in E3, and mounted as previously described (Stawicki et al., 2014). Larvae were transferred into a Nunc Lab-Tek chambered glass (ThermoFisher scientific 155379PK) and stabilized under a nylon mesh and two slice anchors (Warner instruments). To avoid inconsistencies in mitotracker retention over time, each larva was individually stained and then immediately imaged. Z-stack images were acquired using an Olympus Fluoview confocal laser scanning microscope (FV1000) with a 60× oil-immersion objective lens and Fluoview software (FV10-ASW 4.2). A maximum-intensity projection of Z-stacks that covered the full depth of the neuromast hair cells was used for ImageJ analysis. For each neuromast, the maximum-intensity projection was first inverted, sharpened, and then automatically thresholded before the ‘analyze particles’ function was used to measure the average mitochondrial circularity. Orthogonal view and 3D rendering were obtained using Imaris software (Bitplane).

Neomycin-Texas Red

Request a detailed protocol

To examine autophagy, Conjugation of Texas Red-X-succinimidyl ester to neomycin sulfate hydrate (Sigma-Aldrich) was done following the previously described protocol (Stawicki et al., 2014). To make a final concentration of 10 μM of neomycin-Texas Red, neomycin sulfate hydrate was dissolved in nuclease-free water to 50% of the final solution volume. 0.5 M K2CO3 at pH 9.0 was added at 17.6% final volume. Texas Red-X-succinimidyl ester (Life Technologies) was dissolved in dimethylformamide at 2.5 mM and was added at 12% final volume. The volume of the mixture was brought to 100% with deionized water and the solution incubated overnight at 4°C to allow the conjugation reaction to go to completion. Larvae were anesthetized in 0.02% tricaine (Sigma-Aldrich) in E3 and transferred to the imaging chamber. Z-stack images of each neuromast were acquired at baseline (prior to Neo-TR exposure) for 2.5 min at 30 s intervals. Following the addition of neomycin-Texas Red to the larva, images were acquired at 30 s intervals for 1 hr. Hair cells within each neuromast that were visually accessible with no significant physical overlap with neighboring hair cells were selected for analysis. Neomycin-Texas Red puncta were counted manually for each hair cell. For the rescue experiment, larvae were treated with 120 μM NBI-31772 for 24 hr starting from 4 dpf. At 5 dpf, larvae were anesthetized in 0.02% tricaine in 133 μM NBI-31772 and transferred to the imaging chamber. Neomycin-Texas Red was added to a final concentration of 120 μM NBI-31772. Z-stack summation projections that included the full depth of the neuromast hair cells were used to quantify fluorescent intensity of neomycin-Texas Red. A region of interest around brn3c:GFP-labaled hair cells was drawn with the free hand tool in ImageJ. Background fluorescent intensity was measured by drawing a ROI away from the neuromast in the same Z-stack summation projection. The CTCF was used to subtract background fluorescence. The CTCF formula was as follows: Integrated Density – (Area of selected cells × Mean fluorescence of background) (McCloy et al., 2014). F0 is the CTCF value at baseline (before the addition of Neomycin-Texas Red), and F is the CTCF value after the addition of Neomycin-Texas Red. The ΔF/F0 was calculated for each time point as follows: the CTCF value after the addition of Neomycin-Texas Red (F)/the CTCF value at baseline (F0). Time lapse movies were made using ImageJ and corrected for xy drift using the ImageJ plugins StackReg and TurboReg (Thévenaz et al., 1998).

Electron microscopy

Request a detailed protocol

Five dpf pappaa+/+ and pappaap170 larvae in a Tubingen long-fin (TLF) background were processed as previously described (Allwardt et al., 2001). Larvae were fixed for 15 min at 4°C using 1% paraformaldehyde, 1.6% glutaraldehyde, 0.15 mm CaCl2, and 3% sucrose in 0.06 M phosphate buffer, pH 7.4. Larvae were then rinsed and post-fixed in osmium tetroxide (2% in phosphate buffer) for 30 min at 4°C then 1.5 hr at room temperature. The larvae were rinsed in phosphate buffer and in maleate buffer (0.05 M, pH 5.9) and then processed in a solution of 2% uranyl acetate in maleate buffer. Larvae were dehydrated in a graded series of ethanol concentrations. Larvae were then incubated in propylene oxide for 20 min, followed by infiltration with Araldite/Epon resin. Lastly, larvae were cured for 72 hr at 60°C. Eighty nanometer thick sections were mounted on slot grids and post-stained with lead citrate and saturated uranyl acetate. Images were acquired with a Philips CM120 scanning transmission electron microscope with a BioSprint 12 series digital camera using AMT Image Capture Software Engine V700 (Advanced Microscopy Techniques). All images were of head neuromasts, particularly IO 1–3 and SO 1–3. Mitochondrial morphology and ER–mitochondria associations were analyzed using ImageJ (Schneider et al., 2012). For mitochondrial morphology, a ROI was manually drawn around each mitochondrion using the free hand tool. The ‘Analyze Particles’ function was used to measure the area, perimeter, and circularity. Mitochondrial interconnectivity was calculated as the area/perimeter (Wiemerslage and Lee, 2016). To measure the inter-organelle distance, the line segment tool was used to draw a straight line connecting the two closest points between the ER and mitochondria. Any ER–mitochondria associations with distances exceeding 100 nm were discarded from the final analysis. The frequency of ER–mitochondria associations was determined by manually counting the number of ER fragments that were within 100 nm of each mitochondrion.

Statistics

All data were analyzed using GraphPad Prism Software 7.0b for statistical significance (GraphPad Software Incorporated, La Jolla, CA, RRID:SCR_002798). Data are presented as the mean ± standard error of the mean (SEM). The statistical tests and sample size for each experiment are stated in the figure legends. Power analysis of sample size using JMP Pro 15.0 software (SAS Institute Inc) revealed that all our experiments had sufficient power to avoid types 1 and 2 errors. The assumption of normality was tested using Shapiro-Wilk’s test. Parametric analyses were performed using a two-tailed unpaired t-test with Welch’s correction, multiple t-tests with a Holm–Sidak correction, or two-way ANOVA with a Holm–Sidak correction. Non-parametric analyses were performed using a two-tailed t-test with Mann–Whitney correction. Significance was set at p<0.05. All data presented are from individual experiments expect for data in Figures 1F, 2D, and 5C. Hair cell survival data collected from multiple experiments were normalized to their respective controls.

Data availability

All data generated or analyzed during this study are included in the manuscript and supporting files. Source data files are provided for Figures 1–5, and Figure 2—figure supplement 1.

References

  28. Book
    1. Eggermont JJ
    (2017)
    Hearing Loss: Causes, Prevention, and Treatment
    Academic Press.
    1. Welihinda AA
    2. Tirasophon W
    3. Kaufman RJ
    (1999)
    The cellular response to protein misfolding in the endoplasmic reticulum
    Gene Expression 7:293–300.

Decision letter

  1. David W Raible
    Reviewing Editor; University of Washington, United States
  2. Didier YR Stainier
    Senior Editor; Max Planck Institute for Heart and Lung Research, Germany
  3. Alison Coffin
    Reviewer; Washington State University, Vancouver, Canada
  4. Katie Kindt
    Reviewer; National Institute on Deafness and Other Communication Disorders, National Institutes of Health, United States

In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses.

Thank you for submitting your article "Pregnancy associated plasma protein-aa regulates endoplasmic reticulum-mitochondria associations" 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 individuals involved in review of your submission have agreed to reveal their identity: Alison Coffin (Reviewer #2); Katie Kindt (Reviewer #3).

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

As the editors have judged that your manuscript is of interest, but as described below that additional experiments are required before it is published, we would like to draw your attention to changes in our revision policy that we have made in response to COVID-19 (https://elifesciences.org/articles/57162). First, because many researchers have temporarily lost access to the labs, we will give authors as much time as they need to submit revised manuscripts. We are also offering, if you choose, to post the manuscript to bioRxiv (if it is not already there) along with this decision letter and a formal designation that the manuscript is "in revision at eLife". Please let us know if you would like to pursue this option. (If your work is more suitable for medRxiv, you will need to post the preprint yourself, as the mechanisms for us to do so are still in development.)

Summary:

This current study builds on previous work from the same group published in eLife. This past work focused on the mechanism that renders lateral line hair cells of pappaa mutants more susceptible to the ototoxin neomycin. This work found that mitochondrial dysfunction was the underlying cause for neomycin susceptibility. This current study expands on the previous work and suggests that not only defects in mitochondria, but also the ER are involved in neomycin susceptibility. The authors use a variety of approaches including TEM, live imaging, pharmacology and RT-qPCR in their present study. Using TEM the authors show that mitochondria – ER associated are more numerous. Furthermore, similar to disrupting mitochondrial calcium pharmacologically, disrupting ER calcium also renders Pappaa-deficient hair cells more susceptible to neomycin. The authors suggest that this ER dysfunction manifests in several ways. They use live imaging to show that in pappaa mutants hair cells are unable to properly package neomycin into autosomes. In addition, via RT-qPCR they show that pappaa mutants have an increased unfolded protein response (UPR). Currently the relationship between all of these pathological issues is unclear, but this work does reveal additional mechanisms that could render loss of Pappaa detrimental to hair cell health. Although the work is well written and presented and statistically sound, it there are several experiments that are needed to strengthen the claims presented in this study.

Essential revisions:

1. Location of TEM micrographs in hair cell

The morphology of organelles can vary based on location within the cell. For example, in hair cells the ER near the nucleus can be distinct from the ER present near the contacts made with efferent neurons or afferent neurons. (https://pubmed.ncbi.nlm.nih.gov/1430341/; https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.2013.267914).

Can the authors indicate what direction the sections (apical-basal or transverse) were taken, where in the hair cells are the sections were taken and how they determined this location?

2. Quantification of mitochondrial fragmentation

It is clear from the TEM cross sections that the mitochondria in hair cells (Figure 3 A) are quite different between pappaa mutants and controls. Whether there are mitochondria or ER networks are present is not apparent from these TEM images. Nor is it entirely clear that the networks are fragmented. The authors use plugins developed for confocal imaging to estimate fragmentation base on circularity and area/perimeter measurement. It is unclear if these measurement translate to hair cells or TEM. In addition to fragmentation in TEM images, the fragmented mitochondria in pappaa mutants are also hard to see in the live, max-projected mitoTimer images.

The mitochondrial networks and fragmentation may be clearer or be better quantified by acquiring super resolution images of hair cells labeled with mitoTracker. In addition, it is possible that the fragmentation may also be visible or more convincing in videos of Z-stacks of mitoTracker label compared to in the max-projected images provided.

3. Examination of hair cell ER morphology

The previous work on Pappaa in zebrafish hair cells focused extensively on the mitochondria while the currently study the shifted the focus to the hair cell ER. While the ER-mito distances are convincing, a more wholistic picture of the amount or distribution of the ER in wildtype and mutant is lacking.

This could be accomplished either using a transgenic line that labels the ER or a KDEL antibody (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4007406/).

4. It qualitatively appears that pappaa mutant hair cells are taking up a greater quantity of fluorescent Neo faster than WT i.e. the fluorescent intensity is greater in more hair cells. Did the authors quantify Neo-TR uptake?

5. Specificity of the pharmacological treatments

The authors perform numerous pharmacological experiments to disturb ER calcium. The authors suggest that their pharmacological manipulations trigger hair cell death due to the alteration in the interplay between ER/mito calcium in hair cells. What concentration of either of these drugs does it take to kill WT hair cells? Dose-response curves comparing WT and mutant would help support the idea that hair-cell death observed is a direct effect of the drugs on hair-cell ER-mitochondria calcium signaling.

Pharmacology is non-cell autonomous and the authors do not present evidence that these compounds specifically impact hair cell ER or mitochondrial calcium. Alternatively, these compound could impact supporting cell ER (https://elifesciences.org/articles/52160) as well as the ER in the innervating afferent or efferent neurons.

More direct evidence show that hair cell mitochondria or ER calcium (measurements using mitoGCaMP such as in the previous study) are impacted by these treatments would make the author's claims more compelling.

6. The disconnect between IGFR1 and results in the current study.

The identify and location of IGFR1 and the IGFBP are still undefined in this system and therefore it remains unclear exactly how IGRR1 or Pappaa impact sensory hair cells. In previous work on pappaa mutants (enhanced startle response, defects in photoreceptor synapse formation, defects in hair cell mitochondria) the role of IGFR1 in these processes was validated. In the current study, the link with IGFR1 is implied throughout.

It is true that the relationship between IGFR1 and Pappa is well characterized and that currently the only known substrates of Pappa are IGFBPs. Despite this work, it is still possible that given the range of phenotypes in pappaa mutants, that Pappa has other protein substrates that have not yet been identified, or other has biological functions unrelated to the IGF system.

To verify IGFR1 in this current study the authors could use NB1-31772 to stimulate IGF1 bioavailability and test whether this rescues either the autophagy or UPR defects in pappaa mutants. Being able to rescue these phenotypes also makes the study more compelling.

7. The authors state that there is not more spontaneous hair cell death in pappaa mutants compared to controls (line 443). Previous work has shown in zebrafish that Usher mutants (cdh23, ush1c, myo7a) also have an early UPR (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4007406/). Similar to pappaa mutants usher mutants have the same # of hair cells compared to controls, indicating no spontaneous hair cell death. But interestingly Usher mutants do have more TUNEL positive hair cells compared to controls, indicating that more hair cells in Usher mutants are in the process of apoptosis. Based on this new finding implicating the UPR response in pappaa mutants, could pappaa mutants, similar to hair cells in Usher mutants be more fragile (neomycin susceptible) as they are more likely to be in the process of apoptosis? A TUNEL label in pappaa mutants could reveal this. In addition, this paper on UPR in Usher mutant hair cells could be a useful paper to add to the discussion.

8. Line 445-451: "Together, these findings suggest that Pappaa may regulate ER-mitochondria associations by promoting ER homeostasis. It is important to note that the ER and mitochondria are engaged in a constant feedback loop." This line of reasoning seems rather circular, considering that the previous study showed Pappaa regulates mitochondrial function. If mitochondrial function is impaired, it seems likely that ER homeostasis would be disrupted as well.

9. Methods: Overall, the methods section needs more detail. All experiments that were not previously performed by the author or the author's lab should have a concise description of what the authors did next to the reference (e.g. fish were imaged under Lab-Tek Chambered Coverglass (Fisher Scientific) where they were immobilized under a nylon mesh and two stainless-steel slice hold-downs (Warner Instruments) per Stawicki et al., 2014) A detailed description of how individual pappaa170 larvae used in experiments were genotyped is needed. A comprehensive description of how mitochondrial circularity was measured using the "mitochondrial morphology" plug-in in ImageJ is needed.

10. Statistics: how did the authors determine the power of the experiments were sufficient to avoid Type I and Type II error?

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

Thank you for resubmitting your work entitled "Pregnancy associated plasma protein-aa regulates endoplasmic reticulum-mitochondria associations" for further consideration by eLife. Your revised article has been reviewed by 2 peer reviewers and the evaluation has been overseen by Didier Stainier as the Senior Editor, and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed before acceptance, as outlined below:

This current study builds on previous work from the same group published in eLife. This past work focused on the mechanism that renders lateral line hair cells of pappaa mutants more susceptible to the ototoxin neomycin. This previous study found that mitochondrial dysfunction was the underlying cause for neomycin susceptibility. This current study expands on the previous work and suggests that not only defects in mitochondria, but also the ER are involved in neomycin susceptibility. The authors use a variety of approaches including TEM, live imaging, pharmacology and RT-qPCR in their present study. Using TEM the authors show that mitochondria – ER associations are more numerous. Furthermore, similar to disrupting mitochondrial calcium pharmacologically, disrupting ER calcium also renders Pappaa-deficient hair cells more susceptible to neomycin. The authors suggest that this ER dysfunction manifests in several ways. They use live imaging to show that in pappaa mutants hair cells are unable to properly package neomycin into autosomes. In addition, via RT-qPCR they show that pappaa mutants have an increased unfolded protein response (UPR). The work is well written and presented and statistically sound.

Overall, this revision addressed a number of specific concerns experimentally, including providing pharmacological evidence that IGF system is involved in the hair cell processes they're examining. However some issues still need to be addressed:

1. More is needed is needed on the locus of action where Pappaa/ IGF signaling regulates the hair cell ER-mitochondria axis. The first author has previously reported pappaa expression to be in supporting cells. However igf1rb expression in neuromasts has also been reported to be predominantly in supporting cells (e.g. https://piotrowskilab.shinyapps.io/neuromast_homeostasis_scrnaseq_2018/). If Pappa/IGF signaling is occurring in supporting cells, the effect on hair cells would be indirect. Resolution of where signaling occurs would be a valuable addition to this study. At the very least, this issue warrants discussion in the manuscript.

2. Currently the relationship between the numerous pathological issues present in pappaa hair cells (mitochondria and ER morphology disruption, lysosomal packaging) is unclear. Does the origin of the deficit rest in the mitochondria, the ER or another organelle? Overall, this work does reveal an additional mechanism (ER calcium) that renders loss of Pappaa detrimental to hair cell health. One caveat of the study is that it relies heavily on pharmacology which impacts not only the ER in hair cells, but also the glial-like supporting cells that surround hair cells, as well as the afferents and efferents that innervate this sensory system.

These points would be worth discussion.

3. The authors indicated in Figure 1A the orientation of the TEM sections and identified which anterior neuromasts. But it was not entirely clear how they determined the location of their organelles relative to neuronal contacts. There was no discussion in the manuscript on how they analyzed the images using the post-synaptic density and synaptic ribbons as landmarks. Additionally, there was no mention of efferent contacts in either the manuscript or the rebuttal.

4. KDEL labeling does not appear to be specifically labeling ER in the images shown in Figure 1F. Instead, the labeling appears completely colocalized with hair cell plasma membrane bound GFP. For comparison, Blanco-Sánchez et al., 2014 showed discrete labeling of membrane within hair cells. Specificity of labeling should be verified if KDEL labeling will be used to measure the amount or distribution of the ER in wildtype and pappaa mutants.

5. The authors did not answer the question of how they determined their experiments were sufficiently powered.

6. The x/z and y/z max intensity images of Mitotracker are better illustrations of what appear to be more fragmented mitochondria. The Imaris 3D reconstructions the authors mentioned in the rebuttal would have been informative, but they did not include them in Figure 3.

7. The P values for non-significance should be reported.

https://doi.org/10.7554/eLife.59687.sa1

Author response

Essential revisions:

1. Location of TEM micrographs in hair cell

The morphology of organelles can vary based on location within the cell. For example, in hair cells the ER near the nucleus can be distinct from the ER present near the contacts made with efferent neurons or afferent neurons. (https://pubmed.ncbi.nlm.nih.gov/1430341/; https://physoc.onlinelibrary.wiley.com/doi/10.1113/jphysiol.2013.267914).

Can the authors indicate what direction the sections (apical-basal or transverse) were taken, where in the hair cells are the sections were taken and how they determined this location?

EM sections (images shown in Figures 1 and 3) were taken along the apical-basal axis through anterior lateral line neuromasts, particularly SO and IO neuromasts. The section orientation is best illustrated by Figure 1B, where the apical stereocilia are at the top of the image and basal is at the bottom. We added more detailed description of the section plane and specific neuromasts to the Results and Methods text, and indicated section plane in Figure 1A. Hair cells were identified based on their central location and darker cytoplasm as noted by (Owens et al., 2007; Behra et al., 2009; Suli et al., 2016). We do not see the stacked ER cistern morphology mentioned in (Chang et al., 1992). This is perhaps because lateral line hair cells more closely resemble type 2 hair cells which lack the perinuclear ER (Song et al., 1995; Nicolson, 2017). We thank the reviewers for drawing our attention to the difference in ER based on its proximity to the synaptic terminal. We re-analyzed our images using the post-synaptic density and synaptic ribbons as landmarks to help us identify pre-synaptic terminals. The data we included in the analysis presented in Figure 1 did not include the “thin near-membrane cistern” mentioned in (Fuchs, 2014).

2. Quantification of mitochondrial fragmentation

It is clear from the TEM cross sections that the mitochondria in hair cells (Figure 3 A) are quite different between pappaa mutants and controls. Whether there are mitochondria or ER networks are present is not apparent from these TEM images. Nor is it entirely clear that the networks are fragmented. The authors use plugins developed for confocal imaging to estimate fragmentation base on circularity and area/perimeter measurement. It is unclear if these measurement translate to hair cells or TEM. In addition to fragmentation in TEM images, the fragmented mitochondria in pappaa mutants are also hard to see in the live, max-projected mitoTimer images.

The mitochondrial networks and fragmentation may be clearer or be better quantified by acquiring super resolution images of hair cells labeled with mitoTracker. In addition, it is possible that the fragmentation may also be visible or more convincing in videos of Z-stacks of mitoTracker label compared to in the max-projected images provided.

We would like to clarify that no plugin was used to analyze the EM images. We used the “free hand” tool to manually draw a ROI around each individual mitochondrion, then used the “analyze particles” function to measure the area, perimeter, and circularity. We understand that mitochondrial fragmentation is difficult to visualize in EM images. Therefore, we changed the text from “Pappaa loss causes mitochondrial fragmentation” to “Pappaa regulates mitochondrial morphology” which we believe is a better interpretation of our results. In addition, to better show the mitochondrial networks, we used Imaris software to make 3D reconstructions of the mitotracker-stained hair cells. In the revised manuscript Figure 3G, we include both the maximum intensity projection and sections in the xz and yz planes. We hope that the difference in mitochondrial morphology between wild type and pappaa mutants is more evident in these updated images.

3. Examination of hair cell ER morphology

The previous work on Pappaa in zebrafish hair cells focused extensively on the mitochondria while the currently study the shifted the focus to the hair cell ER. While the ER-mito distances are convincing, a more wholistic picture of the amount or distribution of the ER in wildtype and mutant is lacking.

This could be accomplished either using a transgenic line that labels the ER or a KDEL antibody (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4007406/).

We thank the reviewers for this suggestion. To analyze the amount of ER, we labeled the ER with KDEL antibody and measured the area occupied by KDEL immunolabeling, a quantification approach previously published by another zebrafish group (Blanco-Sánchez et al., 2014). We found that the percentage of neuromast area occupied by KDEL immunolabeling was similar between pappaa mutant and wild type hair cells, suggesting that the increase in ER-mitochondria associations is not caused by an expansion in ER. These new data are shown in Figure 1F and G.

4. It qualitatively appears that pappaa mutant hair cells are taking up a greater quantity of fluorescent Neo faster than WT i.e. the fluorescent intensity is greater in more hair cells. Did the authors quantify Neo-TR uptake?

We quantified Neo-TR fluorescence intensity by measuring the change in Neo-TR fluorescence over baseline (ΔF/F0) at 2.5, 4.5, and 6.5 minutes after exposure, and found no difference between genotypes suggesting the cells have similar rates of Neo-TR uptake. Additionally, we measured the maximum Neomycin-TR ΔF/F0 and found no difference between wild type and pappaa mutant hair cells, suggesting similar total amounts of neomycin are taken up by wild type and mutant hair cells. These new data are shown in Figure 4D and E.

5. Specificity of the pharmacological treatments

The authors perform numerous pharmacological experiments to disturb ER calcium. The authors suggest that their pharmacological manipulations trigger hair cell death due to the alteration in the interplay between ER/mito calcium in hair cells. What concentration of either of these drugs does it take to kill WT hair cells? Dose-response curves comparing WT and mutant would help support the idea that hair-cell death observed is a direct effect of the drugs on hair-cell ER-mitochondria calcium signaling.

We agree that dose-response curves provide a more reliable measure of drug effect. Given the high cost of Adensophostin A, we chose to focus on Thapsigargin experiments. In the revised manuscript, we include dose-response data for two ranges of Thapsigargin concentrations: lower concentrations with chronic exposure (Figure 2C) and higher concentrations with acute exposure (Figure 2D). Both these ranges span from doses that cause little to no hair cell death in wild type larvae to doses that cause low levels of wildtype cell death. At all doses, the mutants have lower hair cell survival rates than wild type.

Pharmacology is non-cell autonomous and the authors do not present evidence that these compounds specifically impact hair cell ER or mitochondrial calcium. Alternatively, these compound could impact supporting cell ER (https://elifesciences.org/articles/52160) as well as the ER in the innervating afferent or efferent neurons.

More direct evidence show that hair cell mitochondria or ER calcium (measurements using mitoGCaMP such as in the previous study) are impacted by these treatments would make the author's claims more compelling.

We agree that we cannot rule out potential non-cell autonomous effects of Thapsigargin on neighboring cells. However, we specifically chose Thapsigargin because it was in fact previously shown to increase mitochondrial calcium levels (using mitoGCaMP fluorescence measurements) in zebrafish lateral line hair cells (Esterberg et al., 2014). We added the following text in the revised manuscript: “A previous study used a transgenic zebrafish line with a mitochondria-targeted calcium indicator (mitoGCaMP3) to show that Thapsigargin treatment caused increased mitochondrial calcium uptake in lateral line hair cells (Esterberg et al., 2014).”

6. The disconnect between IGFR1 and results in the current study.

The identify and location of IGFR1 and the IGFBP are still undefined in this system and therefore it remains unclear exactly how IGRR1 or Pappaa impact sensory hair cells. In previous work on pappaa mutants (enhanced startle response, defects in photoreceptor synapse formation, defects in hair cell mitochondria) the role of IGFR1 in these processes was validated. In the current study, the link with IGFR1 is implied throughout.

It is true that the relationship between IGFR1 and Pappa is well characterized and that currently the only known substrates of Pappa are IGFBPs. Despite this work, it is still possible that given the range of phenotypes in pappaa mutants, that Pappa has other protein substrates that have not yet been identified, or other has biological functions unrelated to the IGF system.

To verify IGFR1 in this current study the authors could use NB1-31772 to stimulate IGF1 bioavailability and test whether this rescues either the autophagy or UPR defects in pappaa mutants. Being able to rescue these phenotypes also makes the study more compelling.

We include three new experiments in the revised manuscript to determine whether pappaa acts via IGF1 signaling to exert its effects on ER-calcium transfer and autophagy. First, we show that inducing expression of a dominant negative form of the IGF1 receptor increases hair cell sensitivity to Thapsigargin in pappaa mutants (Figure 2E). Second, we found that treatment with NBI-31772 to increase IGF1 bioavailability improves hair cell survival in pappaa mutant larvae treated with Thapsigargin (Figure 2F). Third, we show that treatment with NBI-31772 also rescues the autophagy defects in pappaa mutant hair cells (Figure 4F-G).

7. The authors state that there is not more spontaneous hair cell death in pappaa mutants compared to controls (line 443). Previous work has shown in zebrafish that Usher mutants (cdh23, ush1c, myo7a) also have an early UPR (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4007406/). Similar to pappaa mutants usher mutants have the same # of hair cells compared to controls, indicating no spontaneous hair cell death. But interestingly Usher mutants do have more TUNEL positive hair cells compared to controls, indicating that more hair cells in Usher mutants are in the process of apoptosis. Based on this new finding implicating the UPR response in pappaa mutants, could pappaa mutants, similar to hair cells in Usher mutants be more fragile (neomycin susceptible) as they are more likely to be in the process of apoptosis? A TUNEL label in pappaa mutants could reveal this. In addition, this paper on UPR in Usher mutant hair cells could be a useful paper to add to the discussion.

We thank the reviewers for this suggestion. We used TUNEL labeling and found that pappaa hair cells are not in the process of apoptosis. These new data are shown in Figure 5C. This result, together with our finding that the ER in pappaa mutant hair cells does not appear to undergo expansion (based on EM images and KDEL immunolabeling shown in Figure 1), which is indicative of more advanced ER stress, suggest that pappaa mutant hair cells experience less advanced ER stress compared to Usher mutants which do show ER expansion. We added text discussing the comparison between pappaa and Usher mutants to our discussion (page 8, lines 344-348)

8. Line 445-451: "Together, these findings suggest that Pappaa may regulate ER-mitochondria associations by promoting ER homeostasis. It is important to note that the ER and mitochondria are engaged in a constant feedback loop." This line of reasoning seems rather circular, considering that the previous study showed Pappaa regulates mitochondrial function. If mitochondrial function is impaired, it seems likely that ER homeostasis would be disrupted as well.

We modified the text in that paragraph to more clearly describe our reasoning and interpretation of our results. Although it is difficult to identify the primary defect when the ER-mitochondrial axis is disrupted, our work has further defined the cellular processes affected by loss of Pappaa. The new text (page 8-9, lines 355-373) reads as follows:

“Together, these findings suggest that Pappaa may regulate ER-mitochondria associations by promoting ER homeostasis. pappaap170 mutants display all the telltale signs of early ER stress, including activation of the adaptive UPR branch, heightened mitochondrial bioenergetics evident by increased calcium load, ROS, and hyperpolarization (Alassaf et al., 2019), and lack of spontaneous apoptosis (Figure 5C). […] Furthermore, given the novelty of Pappaa’s role in regulating ER-mitochondria associations, it will be interesting to identify the downstream molecular targets of Pappaa-mediated IGF1 signaling within the ER-mitochondria tethering complex.”

9. Methods: Overall, the methods section needs more detail. All experiments that were not previously performed by the author or the author's lab should have a concise description of what the authors did next to the reference (e.g. fish were imaged under Lab-Tek Chambered Coverglass (Fisher Scientific) where they were immobilized under a nylon mesh and two stainless-steel slice hold-downs (Warner Instruments) per Stawicki et al., 2014) A detailed description of how individual pappaa170 larvae used in experiments were genotyped is needed. A comprehensive description of how mitochondrial circularity was measured using the "mitochondrial morphology" plug-in in ImageJ is needed.

We revised our manuscript to add more detail throughout the methods section.

10. Statistics: how did the authors determine the power of the experiments were sufficient to avoid Type I and Type II error?

In general, we use the number of larvae as the N for statistical analyses, however, we analyzed at least 3 neuromasts per larva. In the revised manuscript, we include detailed numbers of neuromasts and larvae in the figure legends. All pharmacological experiments included a minimum of 24 neuromasts from 8 larvae per group which we believe to be sufficient to avoid type I and II error. For live imaging experiments, we were somewhat limited in the number of larvae we could test, because in order to maintain consistency of conditions among samples, we imaged all experimental and control groups for each experiment on the same day. The live imaging of Neo-TR autophagy experiment included 22 hair cells from 4 larvae per genotype, and the mitotracker experiment included 8 larvae per genotype. These numbers are in line with other zebrafish lateral line publications.

Alassaf M, Daykin EC, Mathiaparanam J, Wolman MA (2019) Pregnancy-associated plasma protein-aa supports hair cell survival by regulating mitochondrial function. eLife 8.

Behra M, Bradsher J, Sougrat R, Gallardo V, Allende ML, Burgess SM (2009) Phoenix is required for mechanosensory hair cell regeneration in the zebrafish lateral line. PLoS Genet 5:e1000455.

Blanco-Sánchez B, Clément A, Fierro J, Washbourne P, Westerfield M (2014) Complexes of Usher proteins preassemble at the endoplasmic reticulum and are required for trafficking and ER homeostasis. Dis Model Mech 7:547-559.

Chang JS, Popper AN, Saidel WM (1992) Heterogeneity of sensory hair cells in a fish ear. J Comp Neurol 324:621-640.

Esterberg R, Hailey DW, Rubel EW, Raible DW (2014) ER-mitochondrial calcium flow underlies vulnerability of mechanosensory hair cells to damage. J Neurosci 34:9703-9719.

Fuchs PA (2014) A 'calcium capacitor' shapes cholinergic inhibition of cochlear hair cells. J Physiol 592:3393-3401.

Kjaer-Sorensen K, Engholm DH, Kamei H, Morch MG, Kristensen AO, Zhou J, Conover CA, Duan C, Oxvig C (2013) Pregnancy-associated plasma protein A (PAPP-A) modulates the early developmental rate in zebrafish independently of its proteolytic activity. J Biol Chem 288:9982-9992.

Miller AH, Howe HB, Krause BM, Friedle SA, Banks MI, Perkins BD, Wolman MA (2018) Pregnancy-Associated Plasma Protein-aa Regulates Photoreceptor Synaptic Development to Mediate Visually Guided Behavior. J Neurosci 38:5220-5236.

Nicolson T (2017) The genetics of hair-cell function in zebrafish. J Neurogenet 31:102-112.

Owens KN, Cunningham DE, MacDonald G, Rubel EW, Raible DW, Pujol R (2007) Ultrastructural analysis of aminoglycoside-induced hair cell death in the zebrafish lateral line reveals an early mitochondrial response. J Comp Neurol 502:522-543.

Song J, Yan HY, Popper AN (1995) Damage and recovery of hair cells in fish canal (but not superficial) neuromasts after gentamicin exposure. Hear Res 91:63-71.

Suli A, Pujol R, Cunningham DE, Hailey DW, Prendergast A, Rubel EW, Raible DW (2016) Innervation regulates synaptic ribbons in lateral line mechanosensory hair cells. J Cell Sci 129:2250-2260.

Wilczak N, de Vos RA, De Keyser J (2003) Free insulin-like growth factor (IGF)-I and IGF binding proteins 2, 5, and 6 in spinal motor neurons in amyotrophic lateral sclerosis. Lancet 361:1007-1011.

Wolman MA, Jain RA, Marsden KC, Bell H, Skinner J, Hayer KE, Hogenesch JB, Granato M (2015) A genome-wide screen identifies PAPP-AA-mediated IGFR signaling as a novel regulator of habituation learning. Neuron 85:1200-1211.

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

[…] Overall, this revision addressed a number of specific concerns experimentally, including providing pharmacological evidence that IGF system is involved in the hair cell processes they're examining. However some issues still need to be addressed:

1. More is needed is needed on the locus of action where Pappaa/ IGF signaling regulates the hair cell ER-mitochondria axis. The first author has previously reported pappaa expression to be in supporting cells. However igf1rb expression in neuromasts has also been reported to be predominantly in supporting cells (e.g. https://piotrowskilab.shinyapps.io/neuromast_homeostasis_scrnaseq_2018/). If Pappa/IGF signaling is occurring in supporting cells, the effect on hair cells would be indirect. Resolution of where signaling occurs would be a valuable addition to this study. At the very least, this issue warrants discussion in the manuscript.

We added an experiment to ask where Pappaa/IGF signaling occurs, using an antibody against phosphorylated IGF1R (pIGF1R), which recognizes the IGF1R autophosphorylation event that occurs when the receptor is activated by IGF1 binding. This antibody has previously been validated using the IGF1R inhibitor NVP (Chablais and Jazwinska, 2010). We quantified the fluorescent intensity of pIGF1R in neuromasts and found that loss of Pappaa caused a reduction in pIGF1R fluorescence in both the hair cells and the support cells. The data are reported in Figure 2—figure supplement 1. The results and discussion of potential direct and indirect effects are added to the manuscript text on p. 6, lines 247-265. The detailed methods are reported in the Methods section p. 13, lines 521-528.

2. Currently the relationship between the numerous pathological issues present in pappaa hair cells (mitochondria and ER morphology disruption, lysosomal packaging) is unclear. Does the origin of the deficit rest in the mitochondria, the ER or another organelle? Overall, this work does reveal an additional mechanism (ER calcium) that renders loss of Pappaa detrimental to hair cell health. One caveat of the study is that it relies heavily on pharmacology which impacts not only the ER in hair cells, but also the glial-like supporting cells that surround hair cells, as well as the afferents and efferents that innervate this sensory system.

These points would be worth discussion.

We have added more discussion of these caveats and questions to the manuscript text, in the Conclusion section, p. 9-10, lines 404-436.

3. The authors indicated in Figure 1A the orientation of the TEM sections and identified which anterior neuromasts. But it was not entirely clear how they determined the location of their organelles relative to neuronal contacts. There was no discussion in the manuscript on how they analyzed the images using the post-synaptic density and synaptic ribbons as landmarks. Additionally, there was no mention of efferent contacts in either the manuscript or the rebuttal.

We have added a new supplemental figure (Figure 1—figure supplement 1) showing images of the efferent contacts and synaptic ribbons in our EM sections. We added the following text to the manuscript, p. 4, lines 147-153:

“Because the ER at post-synaptic sites adjacent to efferent inputs has a distinct role in buffering high levels of post-synaptic calcium influx, and thus may impact the ER-mitochondria axis differently (Moglie et al., 2018), we did not include ER within 100 nm of the post-synaptic sites in our analysis. […] We also did not include ER in EM sections with presynaptic terminals, identified by synaptic ribbons (Figure 1—figure supplement 1B-B’).”

4. KDEL labeling does not appear to be specifically labeling ER in the images shown in Figure 1F. Instead, the labeling appears completely colocalized with hair cell plasma membrane bound GFP. For comparison, Blanco-Sánchez et al., 2014 showed discrete labeling of membrane within hair cells. Specificity of labeling should be verified if KDEL labeling will be used to measure the amount or distribution of the ER in wildtype and pappaa mutants.

The nuclei in the macula hair cells shown in the Blanco-Sanchez et al. study do not occupy as large a proportion of the cell as the nuclei in lateral line hair cells, thus allowing the ER in the cytoplasm to be spatially distinguished from the plasma membrane. In lateral line hair cells, the nucleus is large and the cytoplasm surrounding the nucleus is contained in a thin layer between the nucleus and plasma membrane. The previous image we showed in Figure 1F was of a single plane through the part of the cell containing the nucleus, and the resolution of our light microscopy was not sufficiently high to distinguish the thin layer of cytoplasm from the plasma membrane. We replaced that image with a maximum projection of several z-planes, including regions of the cytoplasm apical to the nucleus, where cytoplasmic ER labeling is distinguishable from the plasma membrane. The KDEL antibody was shown by Blanco-Sanchez (2018) to specifically label the ER by comparing to a transgenic line Tg[myo6b:kdel-crimson]b1319 that labels the ER.

To complement the KDEL antibody labeling, we used another measure to assess the amount of ER present. We quantified the length of ER profiles in our EM sections, reasoning that increased abundance of ER in the cells may be reflected in longer tubule profiles. We found no difference in ER tubule profile length in mutants versus wild type hair cells. We also did not find a difference in the total number of ER tubules between mutants and wild type. We report these data in the new Figure 1H and have added the results to the manuscript text on p. 4, lines 169-173.

5. The authors did not answer the question of how they determined their experiments were sufficiently powered.

We used JMP Pro 15.0 software to perform power analyses for our experiments. Power analysis on the number of neuromasts, hair cells, or mitochondria we used for each experiment showed that all our experiments were sufficiently powered to avoid type 1 and 2 errors. We include this information in our Methods section, p. 16, lines 647-649, and added it to our Transparent Reporting form.

6. The x/z and y/z max intensity images of Mitotracker are better illustrations of what appear to be more fragmented mitochondria. The Imaris 3D reconstructions the authors mentioned in the rebuttal would have been informative, but they did not include them in Figure 3.

We now include videos of the 3D rendering of wild type and pappaa mutant mitotracker-labeled neuromasts (Video 1 and 2).

7. The P values for non-significance should be reported.

We included the p-values in the manuscript and the source data files

Blanco-Sánchez B, Clément A, Fierro J, Stednitz S, Phillips JB, Wegner J, Panlilio JM, Peirce JL, Washbourne P, Westerfield M (2018) Grxcr1 Promotes Hair Bundle Development by Destabilizing the Physical Interaction between Harmonin and Sans Usher Syndrome Proteins. Cell Rep 25:1281-1291.e1284.

Chablais F, Jazwinska A (2010) IGF signaling between blastema and wound epidermis is required for fin regeneration. Development 137:871-879.

Moglie MJ, Fuchs PA, Elgoyhen AB, Goutman JD (2018) Compartmentalization of antagonistic Ca. Proc Natl Acad Sci U S A 115:E2095-E2104.

https://doi.org/10.7554/eLife.59687.sa2

Article and author information

Author details

  1. Mroj Alassaf

    1. Department of Integrative Biology, University of Wisconsin, Madison, United States
    2. Department of Neuroscience, University of Wisconsin, Madison, United States
    3. Neuroscience Training Program, University of Wisconsin, Madison, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Writing - original draft
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6277-9417

  2. Mary C Halloran

    1. Department of Integrative Biology, University of Wisconsin, Madison, United States
    2. Department of Neuroscience, University of Wisconsin, Madison, United States
    3. Neuroscience Training Program, University of Wisconsin, Madison, United States
    Contribution
    Resources, Supervision, Funding acquisition, Project administration, Writing - review and editing
    For correspondence
    mchalloran@wisc.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6086-5928

Funding

Ministry of Education – Kingdom of Saudi Arabi (Graduate Student Fellowship)

  • Mroj Alassaf

University of Wisconsin-Madison (UW internal funds)

  • Mary C Halloran

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Acknowledgements

The authors would like to thank Benjamin August (UW School of Medicine and Public Health electron microscope facility) for his technical expertise, Dr. Yevgenya Grinblat (University of Wisconsin-Department of Integrative Biology) for the use of the RT-qPCR cycler, Emily Daykin for her assistance with data acquisition, and Daniel North for fish care and maintenance.

Ethics

Animal experimentation: This study was performed in accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. Animals were handled according to approved institutional animal care and use committee (IACUC) protocols (L005704) of the University of Wisconsin.

Senior Editor

  1. Didier YR Stainier, Max Planck Institute for Heart and Lung Research, Germany

Reviewing Editor

  1. David W Raible, University of Washington, United States

Reviewers

  1. Alison Coffin, Washington State University, Vancouver, Canada
  2. Katie Kindt, National Institute on Deafness and Other Communication Disorders, National Institutes of Health, United States

Publication history

  1. Received: June 9, 2020
  2. Accepted: March 23, 2021
  3. Accepted Manuscript published: March 24, 2021 (version 1)
  4. Version of Record published: April 6, 2021 (version 2)

Copyright

© 2021, Alassaf and Halloran

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 797
    Page views
  • 139
    Downloads
  • 1
    Citations

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

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Mroj Alassaf
  2. Mary C Halloran
(2021)
Pregnancy-associated plasma protein-aa regulates endoplasmic reticulum–mitochondria associations
eLife 10:e59687.
https://doi.org/10.7554/eLife.59687

Categories and tags

Research organism

Further reading

    1. Cell Biology
    2. Neuroscience

    Pregnancy-associated plasma protein-aa supports hair cell survival by regulating mitochondrial function

    Mroj Alassaf, Emily C Daykin ... Marc A Wolman
    Research Article Updated

    To support cell survival, mitochondria must balance energy production with oxidative stress. Inner ear hair cells are particularly vulnerable to oxidative stress; thus require tight mitochondrial regulation. We identified a novel molecular regulator of the hair cells’ mitochondria and survival: Pregnancy-associated plasma protein-aa (Pappaa). Hair cells in zebrafish pappaa mutants exhibit mitochondrial defects, including elevated mitochondrial calcium, transmembrane potential, and reactive oxygen species (ROS) production and reduced antioxidant expression. In pappaa mutants, hair cell death is enhanced by stimulation of mitochondrial calcium or ROS production and suppressed by a mitochondrial ROS scavenger. As a secreted metalloprotease, Pappaa stimulates extracellular insulin-like growth factor 1 (IGF1) bioavailability. We found that the pappaa mutants’ enhanced hair cell loss can be suppressed by stimulation of IGF1 availability and that Pappaa-IGF1 signaling acts post-developmentally to support hair cell survival. These results reveal Pappaa as an extracellular regulator of hair cell survival and essential mitochondrial function.

    1. Cell Biology
    2. Neuroscience

    Interaction between Teneurin-2 and microtubules via EB proteins provides a platform for GABAA receptor exocytosis

    Sotaro Ichinose, Yoshihiro Susuki ... Hirohide Iwasaki
    Research Article

    Neurons form dense neural circuits by connecting to each other via synapses and exchange information through synaptic receptors to sustain brain activities. Excitatory postsynapses form and mature on spines composed predominantly of actin, while inhibitory synapses are formed directly on the shafts of dendrites where both actin and microtubules (MTs) are present. Thus, it is the accumulation of specific proteins that characterizes inhibitory synapses. In this study, we explored the mechanisms that enable efficient protein accumulation at inhibitory postsynapse. We found that some inhibitory synapses function to recruit the plus end of MTs. One of the synaptic organizers, Teneurin-2 (TEN2), tends to localize to such MT-rich synapses and recruits MTs to inhibitory postsynapses via interaction with MT plus-end tracking proteins EBs. This recruitment mechanism provides a platform for the exocytosis of GABAA receptors. These regulatory mechanisms could lead to a better understanding of the pathogenesis of disorders such as schizophrenia and autism, which are caused by excitatory/inhibitory (E/I) imbalances during synaptogenesis.

    1. Cell Biology
    2. Chromosomes and Gene Expression

    Cohesin-independent STAG proteins interact with RNA and R-loops and promote complex loading

    Hayley Porter, Yang Li ... Suzana Hadjur
    Research Article Updated

    Most studies of cohesin function consider the Stromalin Antigen (STAG/SA) proteins as core complex members given their ubiquitous interaction with the cohesin ring. Here, we provide functional data to support the notion that the SA subunit is not a mere passenger in this structure, but instead plays a key role in the localization of cohesin to diverse biological processes and promotes loading of the complex at these sites. We show that in cells acutely depleted for RAD21, SA proteins remain bound to chromatin, cluster in 3D and interact with CTCF, as well as with a wide range of RNA binding proteins involved in multiple RNA processing mechanisms. Accordingly, SA proteins interact with RNA, and R-loops, even in the absence of cohesin. Our results place SA1 on chromatin upstream of the cohesin ring and reveal a role for SA1 in cohesin loading which is independent of NIPBL, the canonical cohesin loader. We propose that SA1 takes advantage of structural R-loop platforms to link cohesin loading and chromatin structure with diverse functions. Since SA proteins are pan-cancer targets, and R-loops play an increasingly prevalent role in cancer biology, our results have important implications for the mechanistic understanding of SA proteins in cancer and disease.