R E S E A R CH AR T I C L E

A profile of auditory-responsive neurons in the larval zebrafishbrain

Gilles Vanwalleghem1 | Lucy A. Heap1 | Ethan K. Scott1,2

1School of Biomedical Sciences, The

University of Queensland, St Lucia, QLD,

Australia

2The Queensland Brain Institute, The

University of Queensland, St Lucia, QLD,

Australia

Correspondence

Ethan K. Scott, School of Biomedical

Sciences, The University of Queensland, St

Lucia, QLD 4072, Australia.

Email: ethan.scott@uq.edu.au; Gilles

Vanwalleghem, School of Biomedical

Sciences, The University of Queensland, St

Lucia, QLD 4072, Australia.

Email: g.vanwalleghem@uq.edu.au

Funding information

Grant sponsor: NHMRC Project Grant,

Grant number: APP1066887; Grant

sponsor: ARC Future Fellowship, Grant

number: FT110100887; Grant sponsor: A

Simons Foundation Explorer Award, Grant

number: 336331; Grant sponsor: Two ARC

Discovery Project Grants, Grant numbers:

DP140102036, DP110103612; Grant

sponsor: ARC LIEF, Grant number:

LE130100078.

AbstractMany features of auditory processing are conserved among vertebrates, but the degree to which

these pathways are established at early stages is not well explored. In this study, we have

observed single cell activity throughout the brains of larval zebrafish with the goal of identifying

the cellular responses, brain regions, and brain-wide pathways through which these larvae perceive

and process auditory stimuli. Using GCaMP and selective plane illumination microscopy, we find

strong responses to auditory tones ranging from 100 Hz to 400 Hz. We also identify different cat-

egories of auditory neuron with distinct frequency response profiles. Auditory responses occur in

the medial octavolateral nucleus, the torus semicircularis, the medial hindbrain, and the thalamus,

and the flow of information among these regions resembles the pathways described in adult fish

and mammals. The details of these patterns, however, indicate that auditory processing is still rudi-

mentary in larvae. The range of frequencies detected is small, and while different neurons have

distinct response profiles, most are sensitive to multiple frequencies, and distinct categories show

substantial overlap in their responses. Likewise, while there are signs of nascent spatial representa-

tions of frequency in the larval brain, this only faintly resembles the clear tonotopy seen in adult

fish and mammals. Overall, our results show that many fundamental properties of the auditory sys-

tem are established early in development, and suggest that zebrafish will provide a good model in

which to study the development and refinement of these pathways.

K E YWORD S

Auditory Pathways, Cochlear Nucleus, Fluorescence, RRID: AB_2315112, RRID: AB_2534077,

RRID: SCR_002234, RRID: SCR_001622, RRID: SCR_007198, RRID: SCR_002285, RRID:

SCR_002798, RRID: SCR_001905, Zebrafish

1 | INTRODUCTION

The auditory systems of fish draw information from a range of sensory

structures and neural pathways, and these vary from species to species

and across development. Fish sense sound waves as particle motion

via the otoliths in the inner ear. Amongst “hearing specialists” such as

zebrafish, the Weberian ossicles conduct high frequency swim bladder

vibrations to the inner ear, expanding the hearing range (Grande &

Young, 2004; Higgs, Rollo, Souza, & Popper, 2003; Popper & Fay,

1999). In zebrafish larvae, audition of frequencies above 100 Hz is

likely mediated solely by the saccular otoliths (Popper & Fay, 1993;

Yao, DeSmidt, Tekin, Liu, & Lu, 2016), as the utricular otoliths are

involved exclusively with vestibular perception (Riley & Moorman,

2000), and the Weberian ossicles are neither fully developed nor

functional in larvae (Grande & Young, 2004; Higgs et al., 2003). The

same holds true for the lagenar otolith, which may be involved in audi-

tion, but only develops 2 weeks post fertilization (Bever & Fekete,

2002; Riley & Moorman, 2000). Another potential source of auditory

input is the lateral line, which consists of hair cells similar to the ones

of the inner ear, but organized into neuromasts (Kalmijn, 1988; North-

cutt, 1981). In zebrafish larvae, lateral line afferents respond to

frequencies up to 100 Hz, but it is unclear whether or how this infor-

mation interacts with the rest of the auditory system (Levi, Akanyeti,

Ballo, & Liao, 2015).

Anatomical studies in adult teleost fish have shown auditory path-

ways similar to those in mammals. Auditory information from the VIIIth

nerve flows through the octavolateralis nuclei in the hindbrain to the

torus semicircularis (inferior colliculus in mammals) in the midbrain, and

J Comp Neurol. 2017;1–13. wileyonlinelibrary.com/journal/cne VC 2017Wiley Periodicals, Inc. | 1

Received: 1 March 2017 | Revised: 26 May 2017 | Accepted: 29 May 2017

DOI: 10.1002/cne.24258

The Journal ofComparative Neurology

then to the thalamus (Fay & Edds-Walton, 2008). It is unknown

whether this pathway is developed in larval zebrafish, as most studies

of hearing in zebrafish have looked at juveniles or adults (Cervi, Poling,

& Higgs, 2012; Higgs et al., 2003; Higgs, Souza, Wilkins, Presson, &

Popper, 2002; Mueller, 2012; Wang et al., 2015). The few studies

assessing larval hearing have used microphonic potentials to gauge

responses in the ears’ hair cells (Lu & DeSmidt, 2013; Rohmann, Tripp,

Genova, & Bass, 2014; Yao et al., 2016), and this work has shown

responses to tones ranging from 20 Hz to 400 Hz as early as 3 days

postfertilization (dpf) (Yao et al., 2016). By 5–6 dpf, larvae startle to

tones up to 1,000 Hz, although these responses require extremely

strong stimuli (Bhandiwad, Zeddies, Raible, Rubel, & Sisneros, 2013).

Strong vibrations may also activate the semicircular canals hair cells,

but these are not functional in 6 dpf larvae (Beck, Gilland, Tank, &

Baker, 2004).

With its transparency, and given the rise of light-based tools for

observing and manipulating neural activity, the larval zebrafish has

emerged as an important model for dissecting functional circuits in the

brain. So-called optophysiology, in which fluorescent indicators of

activity are monitored throughout the brain, has enabled the study of

the cell types and circuits involved in processing water flow sensation

(Thompson, Vanwalleghem, Heap, & Scott, 2016), proprioception

(Bohm et al., 2016; Fidelin et al., 2015), somatosensation (Douglass,

Kraves, Deisseroth, Schier, & Engert, 2008), and especially vision

(Bianco & Engert, 2015; Bianco, Kampff, & Engert, 2011; Del Bene

et al., 2010; Temizer, Donovan, Baier, & Semmelhack, 2015; Thompson

& Scott, 2016; Thompson et al., 2016; Vladimirov et al., 2014). Larvae

respond to auditory stimuli (Bhandiwad et al., 2013; Lu & DeSmidt,

2013; Rohmann et al., 2014; Yao et al., 2016), but aside from Mauthner

cell driven auditory escape responses (Mu, Li, Zhang, & Du, 2012; Tani-

moto, Ota, Horikawa, & Oda, 2009), little is known about their auditory

processing, and brain-wide optophysiology has not yet been applied to

this modality.

Here, we use a custom-built selective plane illumination micro-

scope (SPIM) to image the calcium indicator GCaMP6f throughout the

brain of 6 dpf larval zebrafish (Chen et al., 2013) while presenting pure

tones as auditory stimuli. We find responses to frequencies up to 800

Hz, but strong consistent responses only occur from 100 Hz to 400

Hz. A majority of the responsive neurons are located in the medial

octavolateralis nuclei (MON) and the medio-dorsal hindbrain, but

sparse responses take place in the thalamus and the torus semicircula-

ris. Using Granger causality, a statistical method to infer causality from

time series, we suggest that information flows through these regions in

a manner similar to what has been described in mammals.

2 | MATERIAL AND METHODS

Zebrafish (Danio rerio) larvae, of either sex, carrying the transgene

elavl3:H2B-GCaMP6f (Chen et al., 2013) were maintained at 28.58C on

a 14 hr ON/10 hr OFF light cycle. Adult fish were maintained, fed, and

mated as previously described by (Westerfield, 2000). All experiments

were carried out in nacre mutant elavl3:H2B:GCaMP6f larvae of the TL

strain (Chen et al., 2013). Larvae at 6 dpf were immobilized in 2% low

melting point agarose (Progen Biosciences, Australia) and imaged at 5

Hz on a custom-built SPIM (Thompson et al., 2016). The microscope

set-up was isolated from vibrations on a micro-g lab table (TMC, USA,

#63–534). In each larva, horizontal planes were imaged in either the

dorso-ventral or ventro-dorsal direction, at 10 mm increments from the

dorsal-most neurons in the brain to the deepest brain region that could

be clearly imaged using SPIM. For most larvae, this resulted in a stack

of images spanning roughly 250 mm dorso-ventrally, and capturing the

entire rostro-caudal and lateral extents of the brain. This means that

most of the brain was robustly sampled, but that some of the deepest

regions (composing the ventral-most 50 mm, approximately) may have

been missed in some larvae, and may therefore be underrepresented in

our dataset. All image acquisition and stimulus presentation was con-

trolled by lManager software (Edelstein, Amodaj, Hoover, Vale, &

Stuurman, 2010).

For all experiments, audio speakers (Logitech Z213) were posi-

tioned 10cm from the animal on both sides (left and right) and slightly

elevated to have a clear line of sight to the imaging chamber. Pure

tones of frequencies ranging from 100 Hz to 800 Hz were played for

1 s, with 5 s in between stimuli. The sound intensity was measured at

70 dB in air at the larva’s position, and was chosen to avoid acoustic

startle responses. The background level of noise was roughly 40 dB.

Each presentation of the stimulus train involved playing these eight fre-

quencies three times, in ascending, descending, and random order.

Since larval zebrafish perceive sound as particle motion, we measured

motion using an accelerometer (PCB Piezotronics, Q353B34) with a

sensitivity of 98.7 mV/g, attached to the imaging chamber. The signal

was acquired using National Instruments SignalExpress, a vibration

input module (NI-9234), and a DAQ (cDaq-9171). The data were proc-

essed using Matlab. The measured accelerations averaged 0.023 m/s2,

but the measured particle motion for 800 Hz was three times higher

than the other frequencies (0.086 m/s2) (Sup. table 1). These intensities

are in line with what has been used in previous studies (Bhandiwad

et al., 2013). One limitation of our measurements was the impossibility

of embedding the accelerometer in agarose and measuring how

embedding may affect the perceived stimulus. Previous studies have

shown that agarose embedding appears to lower the response thresh-

old of Rana catesbeiana tadpoles (Simmons & Flores, 2012).

For phosphorylated Extracellular signal-Regulated Kinase (pERK)

staining, animals were fixed overnight in ice-cold 4% paraformaldehyde

in PBS immediately after presentation of a 10-min long auditory stimu-

lus train. After washing four times in PBS with 0.3% Triton (PBS-T)

animals were incubated in 150 mM Tris-HCl (pH9.0) for 5 min at 238C,

followed by 15 min at 658C. Larvae were washed in PBS-T again (2 3

10 min) before being incubated in 0.05% Trypsin-EDTA (Thermo Fisher

#25300054) on ice for 45 min. Animals were again washed in PBS-T

(3 3 15 minute washes) and were blocked in 5% goat serum, 1% BSA,

1% DMSO in PBS-T for 1 hr at room temperature. pERK was targeted

using Cell Signaling Technology’s Phospho-p44/42 MAPK (Erk1/2)

(Thr202/Tyr204) (D13.14.4E) (#4370, RRID: AB_2315112) Rabbit

primary mAb. This antibody was generated against a synthetic

2 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

phosphopeptide and detects single or dual phosphorylated Erk1/2

(Filosa, Barker, Dal Maschio, & Baier, 2016; Randlett et al., 2015). Pri-

mary antibodies were added to the blocking solution at a concentration

of 1:500 for 72 hr at 48C. Larvae were washed in PBS-T (3 3 15 min)

and were incubated in secondary antibody coupled to AlexaFluor 546

(A-11010, RRID: AB_2534077) in blocking solution at a concentration

of 1:500 for 72 hr at 48C. Stained larvae were embedded in 2% low

melting point agarose, and confocal stacks of the entire animals were

taken on a ZeissLSM 710 inverted microscope with 3 mm slice inter-

vals. Consistent confocal settings were used for all animals.

Confocal stacks were stitched together using the Pairwise Stitch-

ing plugin for ImageJ (Preibisch, Saalfeld, & Tomancak, 2009). Image

registration of anti-tERK expression was performed against a model of

anti-tERK expression in the nervous system of larval zebrafish. This

was performed with Computational Morphometry Toolkit, RRID:

SCR_002234 using the command string -awr 010203 -T 8 -X 52 -C 8

-G 80 -R 3 -A ‘–accuracy 0.4’ -W ‘–accuracy 1.6’. Separately, experi-

mental and control animals were averaged using a custom-written Mat-

lab (RRID: SCR_001622) script (Randlett et al., 2015), which was then

incorporated into a local version of the Z-Brain Atlas.

Neomycin (Sigma, N6386) treatment involved a 1-hr incubation

with 200 mM neomycin in E3 media, followed by two washes in E3 and

a 1-hr recovery period (Harris et al., 2003). The effectiveness of the

neomycin treatment was assessed using DASPEI staining as performed

by Harris et al. (2003). Specifically, the presence or absence of 12 neu-

romasts (N, IO1–4, SO1–3, O1–2, OP1 and M2) per fish (4 controls

and 3 neomycin treated) was scored (Raible & Kruse, 2000).

In calcium imaging experiments, the aforementioned auditory stim-

ulus train was reused. Stimuli were generated in Audacity, RRID:

SCR_007198. Imaging was performed on a house-built SPIM micro-

scope, as previously described (Thompson & Scott, 2016; Thompson

et al., 2016). The essential components of this microscope are listed in

Table 1.

Motion artifacts caused by slow drift of the image or by spontane-

ous movements by the larva were corrected in Fiji, RRID:

SCR_002285, using “moco” (Dubbs, Guevara, & Yuste, 2016) followed

by a rigid body transformation in StackReg (Thevenaz, Ruttimann, &

Unser, 1998). The eyes and otoliths were manually cropped from the

images before segmentation to avoid artifactual signals. Segmentation

of individual cells was done with the MorphoLibJ morphological seg-

mentation tool (Legland, Arganda-Carreras, & Andrey, 2016). ROIs

smaller than 8 pixels total or larger than 200 pixels were excluded. We

averaged the fluorescence across each ROI, and DF/F was calculated

as previously described (Jia, Rochefort, Chen, & Konnerth, 2011). All

DF/F time series from the 15 fish were then analyzed as follows.

Stepwise linear regressions and location analyses were performed

in Matlab using custom scripts. The stepwise regression fits a linear

model to the data using the Bayesian information criterion to select

the predictors that improve the fit of the model. Predictors were built

for each of the eight frequencies, with a typical GCaMP6f response

occurring for each of the three presentations of the tone. The coeffi-

cient of determination (r2) of the linear regression models was used to

select auditory responsive neurons, and we chose a 0.15 threshold

based on the r2 distribution of our models to allow for conservative fil-

tering of the data. All statistical analyses were performed with Graph-

pad Prism 7.0, RRID: SCR_002798.

The correlation clustering approach (Bianco & Engert, 2015) was

performed in Matlab. From a given trace, we computed its correlation

coefficient with all the other fluorescent traces and merged it with the

highest correlated trace above our threshold of 0.85. This high thresh-

old was required because of the broad response profile of most neu-

rons. We repeated the process until no traces showed correlation

above 0.85 to the merged traces. We then started the process again

from another unmerged trace until no pairs of traces and merged traces

had a correlation above 0.85. Clusters were deemed representative if

they contained at least 10 neurons per fish in at least 5 fish. Only rep-

resentative clusters were considered for further analysis. The coordi-

nates of individual neurons were normalized by registering all larvae

onto each other using Matlab. Principal Component Analysis (PCA) was

used to define the axis of maximal variance in order to investigate pos-

sible tonotopy.

Granger causality was computed with the VARS and tseries pack-

age in R (RRID: SCR_001905), following the Toda-Yamamoto proce-

dure (Toda & Yamamoto, 1995). Briefly, we used both augmented

Dickey–Fuller (ADF) and Kwiatkowski–Phillips–Schmidt–Shin (KPSS)

test to assess for the stationarity of our time series and identify the

TABLE 1 SPIM parts list

Component Product Manufacturer

Beam expander LC1715 f5250 mm Concave Lens Thorlabs

Beam expander LA1708 f5 200 mm Convex Lens Thorlabs

Beam splitter BS PLATE 50 3 50 mm 50R/50T Edmund Optics

Fixed slit 6 3 50 mm slit Custom made

Cylindrical lens LENS CYL 50 3 25 mm 3 75 FL VISNIR Edmund Optics

Illumination objective XLFLUOR4X 0.28NA Olympus

Detection objective XLUMPFLN 20XW 1.0NA Olympus

Camera pco.edge 5.5 PCO

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 3

maximal order of integration (Kwiatkowski, Phillips, Schmidt, & Shin,

1992; Said & Dickey, 1984). We then set up a vector autoregressive

(VAR) model of the maximal order of integration (1 in our case) and

used the Akaike Information criterion to select the optimal lag. We

ensured that there was no serial correlation using the portmanteau test

and performed the Johansen test to ensure the time series were coin-

tegrated (Johansen, 1991). Finally, we computed the pairwise Granger

causality of our time series (Granger, 1969).

All procedures were performed with approval from the University

of Queensland Animal Welfare Unit in accordance with approval

SBMS/305/13/ARC.

3 | RESULTS

3.1 | Auditory responsive regions of the larval

zebrafish brain

As an initial means of identifying auditory responsive brain regions at 6

dpf, we played auditory stimuli, fixed the larvae, and stained for pERK

as an early marker of neuronal activation (Fig. 1) (Randlett et al., 2015).

The tones, ranging from 100 Hz to 800 Hz, were played at 70 dB for

10 min (see Methods) immediately prior to fixation. Auditory stimuli

resulted in an increase of the pERK signals in several brain regions (Fig.

1a’–f’), when compared to control larvae not exposed to auditory stim-

uli (Fig. 1a–f).

We observed clear responses along both vglut2 stripe 1 and vmat2

stripe 1 in the hindbrain (Fig. 1d’, red and blue outlines, respectively)

(Kinkhabwala et al., 2011; Randlett et al., 2015; Wen et al., 2008). This

region is known to receive lateral line, auditory, and somatosensory

input and corresponds to the MON in larval zebrafish (Liao & Haehnel,

2012). We also see an overall increase in the pERK signal in the medial

hindbrain, a region of the brain that is still histologically undifferentiated

in larvae (Fig. 1d’, along the midline) (Mueller & Wullimann, 2016). Con-

sistent with previous calcium imaging observations (Thompson et al.,

2016), we see auditory responses in the deep optic tectum (Fig. 1e’, red

outline). Other areas with auditory responses include a weak and diffuse

labelling in the torus semicircularis (Fig. 1f’, red outline) and a strong

labelling in the pallium (Fig. 1b’, c’, blue outlines). The subpallium (Fig. 1c’

cyan outline) also shows a weak increase in the pERK signal, but most of

the telencephalic signal appears localized in the pallium.

These results demonstrate that an auditory pathway in larval

zebrafish is present by 6 dpf, and that it spans both first order sensory

structures and more central brain regions. In order to gauge how these

broad patterns are manifested in the activity of individual neurons

within these structures, we next performed calcium imaging of auditory

processing using SPIM and the calcium indicator GCaMP6f.

3.2 | Auditory neurons respond strongly up to tones

ranging from 100 Hz to 400 Hz

We presented the same 70 dB, 100–800 Hz tones to immobilized lar-

vae expressing H2B-GCaMP6f pan-neuronally, and imaged calcium

dynamics across the brain using SPIM in order to track the responses

of individual neurons (Fig. 2a, b). The tones were presented three

times, in ascending, descending, and random order of frequencies (Fig.

2d, bottom). Movies were preprocessed using motion corrections and

manual cropping (Dubbs et al., 2016; Thevenaz et al., 1998), and indi-

vidual cells were morphologically segmented into distinct regions of

interest (ROIs, Fig. 2c) (Legland et al., 2016). Following the exclusion of

ROIs not corresponding to neurons (see Methods) the average fluores-

cence of each ROI was measured, and the change of fluorescence from

baseline (DF/F) was calculated at each time point (Fig. 2d) (Jia et al.,

2011; Legland et al., 2016).

To avoid the inclusion of spontaneously active nonauditory neu-

rons, we selected neurons active at least six times using correlation to

an average GCaMP6f spike (threshold of 0.75). Although this could

prevent us from finding truly specific neurons, we did not find any

groups of neurons that showed a significant and representative

response to a single frequency. Our approach yielded roughly 120,000

active neurons across eight animals. It has previously been shown that

the lateral line may contribute to hearing, especially at low frequencies,

in zebrafish larvae (Kalmijn, 1988; Levi et al., 2015; Northcutt, 1981).

To rule out the possibility that the lateral line is responsible for our

observed responses, we treated additional larvae with neomycin to

ablate the lateral line hair cells (Harris et al., 2003). Larvae lacking lat-

eral line hair cells (as demonstrated by a lack of DASPEI labelling of

their neuromasts, Table 2) showed similar responses to controls (see

Fig. 3b). This means that the observed responses result from otolith-

generated auditory signalling from the inner ear hair cells, since these

survive bath application of neomycin (Lombarte, Yan, Popper, Chang, &

Platt, 1993). However, we cannot exclude that the lack of response

from the lateral line may be caused by the embedding of the larvae in

agarose, which could impede the motion of the hair cells. Because

responses were equivalent between control and neomycin treated ani-

mals, we pooled our control data with an additional 130,000 respon-

sive neurons drawn from six neomycin treated larvae.

We next sought to identify, among the roughly 250,000 auditory

responsive neurons (Fig. 2e), those responding consistently to specific

tones. We used a stepwise approach to fit linear regression models to

all our fluorescent traces (see Methods). We then selected fluorescence

traces with an r2 value above 0.15 (15% of the data variance is

explained by the linear regression) for further analysis (Fig. 2f). This

resulted in roughly 70,000 traces corresponding to individual consis-

tently responding auditory neurons (Fig. 2g).

In order to identify categories of neurons with distinct response

profiles, we adapted a clustering approach first used by Bianco and

Engert (2015). Briefly, we clustered highly correlated traces together

and averaged their responses; we repeated the process until no unclus-

tered traces showed correlation coefficient above 0.85 to the merged

traces. This produced five clusters with distinct responses (Fig. 3a).

Although each has unique tuning properties, they also have commonal-

ities. All the clusters show broad tuning from 100 Hz to 300 Hz, and

most also show strong responses up to 400 Hz (Fig. 3c). All clusters

also show similar distributions between neomycin treated and control

larvae (Fig. 3b), especially cluster 1, which given its selective response

4 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

FIGURE 1 Auditory responsive brain regions. pERK signals are shown in magenta, H2B-RFP reference is shown in grey (a,c,e) or green(other subpanels) (Randlett et al., 2015). Images represent averages of 5 experimental and 3 control animals. Each row show horizontalimages from dorsal (First row; �50 mm below the skin), intermediate (Second row; �100 mm below the skin) to ventral (Third row;�150 mm below the skin). (a), (c) and (e) show the major brain regions outlined as follow. (a) Optic tectum in red, Cerebellum in cyan,Habenula in green. (c) Optic tectum in red, Cerebellum in cyan, pretectum in green and pallium in blue. (e) Torus semicircularis in red,pallium in blue, subpallium in cyan, thalamus in green and posterior lateral line ganglion in pink. Larvae were embedded in agarose and

presented with no auditory stimuli (a1,b,c1,d,e1,f)) or with auditory stimuli (a2,b1,c2,d1,e2,f1). Higher magnification images (boxes) areshown in rightmost panels (b,d,f). Stronger signals are seen in larvae exposed to auditory tones in both glutamatergic (red) and vmat2stripe 1 (blue), as well as in the medial hindbrain (along the midline gap) (a1–2,b-b1). D1 shows stronger signals in the optic tectumand pallium (red and blue, respectively). Finally, we see an increase in the pERK signal in torus semicircularis, pallium and subpallium(f1, red, blue, and cyan, respectively). The thin dashed white lines indicate major boundaries between fore-, mid- and hindbrain. Scalebars: 100 mm

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 5

to the 100 Hz stimulus (Fig. 3d), might have been expected to be

driven by the lateral line (Levi et al., 2015). Furthermore, we did not

observe auditory responses in the posterior lateral line ganglion neu-

rons. Overall, these results suggest that the lateral line plays no role in

the auditory perception of 6 dpf zebrafish larvae in the 100–800 Hz

bandwidth and at this moderate sound intensity.

In contrast to prior studies using microphonic potential recordings

(Lu & DeSmidt, 2013; Yao et al., 2016), we observe responses, albeit

weak ones, up to 800 Hz (Fig. 3c). Using linear regression, we esti-

mated the tuning properties of each cluster to the different frequencies

and organized the clusters from low to high frequency preferences

(Fig. 3d). Cluster 1 appears strongly tuned to 100 Hz, cluster 2 is

slightly more tuned to 200 Hz, as is cluster 3. Cluster 4 appears more

tuned to 100 and 300 Hz, whereas cluster 5 appears more tuned to

400 Hz. These are relatively subtle distinctions, however, and all clus-

ters except for cluster 1 have broad tuning from 100 Hz to 400 Hz.

3.3 | Brain-wide distribution of responsive clusters,

and the flow of information through the brain

Next, we explored the ways in which these clusters were represented

spatially across auditory responsive brain regions, and asked whether

FIGURE 2 Functional imaging and image processing workflow. (a) shows a single frame from SPIM imaging of a elavl3:H2B:GCaMP6f 6dpflarva. Scale bar: 50 mm. (b) shows a higher magnification image of the indicated area in (a) with individual cell nuclei visible, and (c) showsthe ROIs after morphological segmentation (see Methods) representing individual neurons within this small region of the brain. (d) showsindividual DF/F traces for the two neurons highlighted (red and green) in (c). The bottom row of (d) indicates the stimulus train presentedwith ascending, descending, and random ordering of the 8 frequencies. (e) shows a raster plot of the color-coded DF/F for the 250,000 neu-rons from the 15 fish in our dataset. The coefficient of determination (r2) values of the stepwise linear regression to our 8 predictors of the250,000 neurons from (e) are shown in (f), with a red dashed line indicating our 0.15 r2 threshold. (g) shows the raster plot for the resulting70,000 neurons with an r2 value above 0.15, color-code is the same as (e)

TABLE 2 DASPEI staining, positive neuromasts

Neuromasts Control (n54) Neomycin treated (n53)

N 4/4 0/3

SO1–3 11/12 0/9

IO1–4 15/16 0/12

OP1 – M2 8/8 0/6

O1 – O2 8/8 0/6

DASPEI labelling of 12 neuromasts (N, IO1–4, SO1–3, O1–2, OP1 andM2) per fish (4 controls and 3 neomycin treated), 1 indicate presence oflabelling, while 0 is the absence of labelling of the neuromast. N5 nasal,IO5 infraorbital, SO5 supraorbital, OP5opercular, O5otic,M5mandibular.

6 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

different frequencies were represented in spatially distinct manners. By

mapping the neurons belonging to our five clusters back onto the cor-

responding locations within the brain (Fig. 4a, b), we identified four

brain regions that contained a majority of the auditory responsive neu-

rons: the torus semicircularis, the thalamus, the medial hindbrain, and

the MON (Fig. 4c). These brain regions were defined based on the Z-

brain atlas as in Fig. 1 (Randlett et al., 2015). Among these, the MON

contained a majority of the consistently auditory-responsive neurons.

Beyond these regions, there was a very sparse distribution of auditory

responses in the rest of the brain (including the tectum and cerebellum)

representing 2.8% of the responding cells. The distributions of clusters

across the brain regions showed that all were widely distributed, with-

out any obvious enrichment of particular clusters in particular regions

(Fig. 4c).

Spatial organization of the auditory responses to different frequen-

cies (tonotopy) is a common feature of auditory processing in mammals

(Bourk, Mielcarz, & Norris, 1981; Echteler, 1985b; Kandler, Clause, &

Noh, 2009; W. Lippe & Rubel, 1985; W. R. Lippe, 1987; Muniak &

Ryugo, 2014). Since the majority of our auditory responses were found

in the MON, and since the corresponding cochlear nuclei in mammals

are tonotopic, we analyzed our responses in this region for tonotopy.

We normalized the xyz coordinates for each fish and used principal

component analysis to find the axes of maximal variance on the pooled

coordinates of all the clusters. The main variance axis (explaining 85%

of variation) that we identified was close to the dorso-ventral axis (108

off the vertical axis, Fig. 4b dashed red line 1). In mapping our five clus-

ters along this axis, we found a significant trend in which lower-

frequency clusters were located more ventrally than higher-frequency

clusters (Fig. 4d). No clear trend emerged from the second component,

which explains a remaining 8% of the variance (rostro-caudal axis,

dashed red line 2), but we see significant differences between the clus-

ters, with cluster 4 more rostral and cluster 3 more caudal (Fig. 4e). The

torus semicircularis is known to be tonotopic in adult carp (Echteler,

1985b), but PCA did not reveal any significant difference in the cluster

position in our larval zebrafish.

Because the brain regions that we have identified are homologous

to regions that respond to auditory stimuli in mammals, we were inter-

ested in whether auditory information flows through these circuits in a

manner similar to how it does in mammals. To address this, we used

Granger causality to infer the directional flow of information among

the different brain regions identified in Fig. 4c (Cadotte, DeMarse, He,

& Ding, 2008; Granger, 1969; Seth, Barrett, & Barnett, 2015). Time

FIGURE 3 Clustering of auditory responsive neurons. (a) shows a raster plot of the color-coded DF/F for neurons composing the fiveclusters obtained after correlation merging. The number of neurons contributing to each cluster is indicated to the left. (b) indicates the

distribution of cells (in %) in each cluster across neomycin treated (grey) and untreated animals (white). The mean DF/F traces (1/2 95%confidence interval (CI), dashed line) for all neurons in each of the five clusters are shown in (c). The bars underneath each graph representthe times during which each pure tone was presented to the larva. (d) shows the tuning curve (normalized estimated coefficients 1/2 95%CI) to the eight frequencies for the neurons in each of our five clusters

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 7

FIGURE 4 Spatial organization of auditory responses, and the flow of auditory information through the brain. (a) shows a maximumintensity projection of the locations of auditory responsive neurons, coded by cluster (colors maintained from Fig. 3, Scale bars: 50 mm), and(b) contains a rotated view of a 3D interpolated model of (a). The dorso-ventral axis is indicated, as well as the two components of the PCAfor the cluster coordinates (red dashed line). Major brain regions are outlined as follow; optic tectum in red, thalamus in green and MON inblue (c) Each of the four major auditory responsive brain regions contains a mix of cells from all clusters. (d) shows the mean score (1/2SEM) along the first component of the PCA (from b) for each cluster within the MON. There are no significant differences among the clus-ters (Kruskal Wallis test with Dunn’s multiple comparison test), but there is a significant linear trend (ANOVA test for Trend, p5 .0076). (e)shows the mean score (1/2 SEM) along the second component of the PCA (from b) for each cluster within the MON. Significant differen-ces are indicated by asterisks (Kruskal Wallis test with Dunn’s multiple comparison test, p values: red versus green50.029; red versusorange50.0414; blue versus pink50.0013 and red versus pink <0.0001)

FIGURE 5 Granger causality among the auditory brain regions. (a) p value (top) and F-statistics (bottom) of the Granger causality computa-tions among the four brain regions outlined in Fig. 4. (b) Summary model of (a), where the width of the arrows is proportional to the F-statof the causality, and only significant (Bonferroni corrected p value of .01/12) links are shown

8 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

series X is said to Granger cause another time series Y if the past of X

contains more information than only the past of Y to predict the future

of Y. The average neural responses of each brain region were tested

for Granger causal relations to the responses in each of the other audi-

tory responsive regions. The resulting significant predictions of Granger

causality show a flow of information from the MON and the hindbrain

to the torus semicircularis (Fig. 5). Granger causality also suggests that

the thalamus is driven by the MON, the hindbrain, and the torus semi-

circularis. There also appear to be weaker flows of information from

the torus semicircularis and thalamus back to the hindbrain and the

MON, which may reflect the existence of feedback loops. Alternatively,

these weak apparent causal links may simply result from the regions’

being driven by a common stimulus. Overall, these results confirm the

expected flow of information among the different brain regions, based

on auditory processing in mammals (Demanez & Demanez, 2003;

Grothe, Pecka, & McAlpine, 2010; Kandler et al., 2009).

3.3 | Stimulus intensity

Finally, we explored how the intensity (volume) of the stimulus is

encoded. Stimulus intensity could be represented by rate coding within

responsive neurons, by the number of neurons responding, or by popu-

lations of neurons with distinct sensitivities. We selected the 400 Hz

tone to approach this issue. Supra-threshold stimuli may trigger startle

responses and not accurately reflect environmental stimuli (Bhandiwad

et al., 2013). As such, our strongest stimulus was at 80 dB, which is

twice the intensity used in the previous experiments. We then

decreased the intensity by factors of two, down to 40 dB, or one six-

teenth of the highest intensity. The stimulus train consisted of three

repetitions, one of decreasing intensity tones, followed by increasing

intensities, and finally randomly ordered intensities. Reliably responsive

auditory neurons were selected as before (Fig. 6a, b).

The data were analysed as for our frequency experiments, and we

obtained two clusters of responses (Fig. 6c). The two clusters show

similar response profiles, with a sharp drop in the strength of neural

response below 60 dB (Fig. 6d). The main difference is a plateau-like

response profile between 80 dB and 60 dB for cluster 1 (Fig. 6d, pur-

ple), while cluster 2 shows a more linear decrease in intensity, with a

comparatively smaller drop of response strength between 60 dB and

50 dB. This difference notwithstanding, the two clusters show broadly

similar tuning, and there is no compelling reason to believe that they

are playing distinct functional roles. Since DF/F is a loose proxy for the

number of spikes (Akerboom et al., 2012; Chen et al., 2013), this is con-

sistent with the idea that the intensity of the sound stimulus is rate

encoded in larval zebrafish. This conclusion is strengthened by the fact

that volume-specific clusters are absent, although it does not rule out

temporal coding that calcium imaging would not reveal.

4 | DISCUSSION

To our knowledge, this is the first investigation of auditory processing

at the whole brain level with cellular resolution in any system, and it

demonstrates the feasibility of using calcium imaging to look at audi-

tory processing across the brain.

FIGURE 6 Neural responses to sound intensity. (a) Provides a raster plot of the DF/F from 160,000 auditory responsive neurons acrossseven fish, and (b) gives the same information for the 50,000 neurons with a r2 value above .15. Raster plots for cluster 1 (top) and 2 (bottom)are shown in (c). Mean DF/F traces (1/2 95% CI) and normalized estimated coefficients (1/2 95% CI) for the responses to the five intensitiesare shown in (d) and (e), respectively, for these two clusters. The number of neurons belonging to each cluster in indicated in (d)

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 9

In mammals, the primary auditory pathway involves information

coming from the VIIIth nerve, which is then processed in the cochlear

nuclei. The cochlear nuclei then project to the inferior colliculus

(homologous to the torus semicircularis in fish) and then the thalamus,

which then projects to the auditory cortex (Kandler et al., 2009; Web-

ster, 1992). We have confirmed that the organization of the auditory

pathway in larval zebrafish is similar to that of mammals. The MON,

where we observe our first and most robust auditory responses, is part

of the rhombic lip, specifically in the lateral glutamatergic stripe that

develops into the octavolateralis nucleus (Liao & Haehnel, 2012; Wulli-

mann et al., 2011). This region expresses atoh1 and ptf1a, which give

rise to excitatory and inhibitory neurons, respectively, in the homolo-

gous cochlear nuclei of mammals (Farago, Awatramani, & Dymecki,

2006; Fujiyama et al., 2009). Auditory information flows through the

octavolateralis nuclei in the hindbrain to the torus semicircularis in the

midbrain, and then to the thalamus in the forebrain. These results con-

firm anatomical studies in adult teleosts (Echteler, 1985a; Fay & Edds-

Walton, 2008; Mueller, 2012; Northcutt, 2006), and reinforce the util-

ity of zebrafish model for studying auditory processing that appears to

be grossly conserved from larval fish through mammals.

Interestingly, pERK staining suggests that there are telencephalic

responses to auditory stimuli, but these were not detected by our cal-

cium imaging. Several possibilities exist to explain these results. Telen-

cephalic responses may reflect a more complex encoding of stimulus

properties, which may not be apparent using calcium imaging. Another

possibility is that the high rate of spontaneous activity in the telen-

cephalon (data not shown) prevents the identification of specific audi-

tory responses using our approach. However, previous studies have

shown that the dorsal part of the telencephalon receives auditory input

in the carp and the goldfish (Echteler, 1985a; Northcutt, 2006). Based

on these observations, we speculate the auditory responsive region of

the pallium we observed may correspond to the future dorsomedial

and dorsocentral divisions.

The lateral line has been implicated in directionality of the auditory

startle response (Mirjany, Preuss, & Faber, 2011), but it is unclear

whether it plays a role in normal audition for stimuli that are sub-

threshold for a startle response. We find that ablating the lateral line

does not affect the auditory processing, at least in the 100–800 Hz

range, and at 70 dB, consistent with previous observations of lateral line

responses to lower frequency, but higher intensity sounds (Levi et al.,

2015; Mirjany et al., 2011). Information from the lateral line, like audi-

tory information, flows through the octavolateral nucleus and the torus

semicircularis (Liao & Haehnel, 2012), so it is possible that these modal-

ities converge on downstream premotor circuitry that mediates behav-

ioral responses to stimuli across a range of frequencies and volumes.

Further imaging studies of water-flow responses mediated by the lateral

line, registered against auditory responses, will be needed to resolve

whether and how these streams of sensory information interact.

The processing of different frequencies is spatially organized in

auditory pathways across the animal kingdom. This tonotopy has been

observed in the cochlear nuclei of mice, cats, and chickens (Bourk et al.,

1981; Lippe & Rubel, 1985; Muniak & Ryugo, 2014). Tonotopy has also

been observed in the torus semicircularis of the carp and in the auditory

system of Drosophila (Echteler, 1985b; Lai, Lo, Dickson, & Chiang, 2012).

We have observed a trend from low to high frequency across the MON,

roughly aligned with the dorso-ventral axis. This gradient is not sharp,

possibly due to the early developmental stage at which we perform our

imaging. It is known that the tonotopic maps are refined with age and

experience, so it is possible that this tonotopy is represented more

clearly in later stages. Furthermore, the hearing range of larval zebrafish

is constrained by the lack of the Weberian ossicles (Higgs et al., 2002;

Higgs et al., 2003; Kandler et al., 2009; Cervi et al., 2012; Wang et al.,

2015). As such, a full tonotopic map may await the formation of the

Weberian ossicles, with the associated expansion of frequencies

detected. More generally, we find that our clusters, while unique, are

broadly tuned and overlap extensively in the frequencies to which they

respond. They are also spread across the responsive brain regions with-

out pronounced specificity from region to cluster. All of these observa-

tions point toward a relatively immature auditory system, possibly

capable more of the rudimentary detection of low frequency sounds

than the fine discrimination of a range of auditory stimuli.

We speculate that the regions we observe in the hindbrain (MON

and medial hindbrain) will develop into the known teleost auditory nuclei.

The region we call the MON based on previous studies (Kinkhabwala

et al., 2011) should develop into the known auditory octaval nuclei,

namely the descending and anterior octaval nuclei (Echteler, 1984,

1985b; Meredith & Butler, 1983; Mueller, 2012; Northcutt, 1981). We

speculate the large medial hindbrain response we observe may also take

part in the development of the aforementioned octaval nuclei; it also

may be involved in the development of the secondary octaval population

(Echteler, 1984; McCormick & Hernandez, 1996). We suspect that the

auditory responsive region of the thalamus we observed develops into

the dorso/central posterior nucleus, a known recipient of auditory infor-

mation (Kirsch, Hofmann, Mogdans, & Bleckmann, 2002; Mueller, 2012).

More developmental studies in larvae and juveniles will be required to

elucidate how these various nuclei form and how tonotopy develops.

One limitation of this study is the use of calcium imaging; calcium

indicators such as GCaMP can only inform us of rate or population

coding in the brain. Subtle temporal coding of information, as takes

place for frequency discrimination in mammals (Frisina, 2001; Micheyl,

Schrater, & Oxenham, 2013), is currently unresolvable with the intrinsi-

cally slow responses of calcium indicators. Nonetheless, we see that

rate coding of the frequency may be present in the MON. Further-

more, we see clear indications of rate coding for the auditory stimulus

intensity, consistent with what is observed in mammals (Moore, 2003;

Sachs & Abbas, 1974). The encoding of the intensity of the stimulus, or

loudness, appears to be linear before reaching a plateau 20 dB above

our weakest stimulus. This saturation of the firing rate is also observed

in mammalian neurons (Moore, 2003; Sachs & Abbas, 1974).

Interestingly, we find that auditory responsive neurons are highly

consistent, responding in nearly 100% of trials to their relevant stimuli.

This stands in contrast to sensory coding in the visual system of larval

zebrafish, where a wide range of response probabilities exist, and

where different neurons show sharply different stimulus tuning (Bianco

& Engert, 2015; Del Bene et al., 2010; Orger, Smear, Anstis, & Baier,

2000; Temizer et al., 2015; Thompson & Scott, 2016; Thompson et al.,

2016). This is likely a function of the depth of sensory discrimination

10 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

across these modalities. Vision is the primary means of hunting for

larval zebrafish (Bianco & Engert, 2015; Bianco et al., 2011; Gahtan,

Tanger, & Baier, 2005; Randlett et al., 2015) and also mediates preda-

tor avoidance (Dunn et al., 2016; Temizer et al., 2015), and behavioral

responses to wide-field motion (Easter & Nicola, 1997; Orger & Baier,

2005; Orger et al., 2000; Roeser & Baier, 2003). Accordingly, larvae

must be able to distinguish among numerous types of visual stimuli.

Audition at this developmental stage may simply allow the larva to per-

ceive generic auditory stimuli with some level of volume perception.

This simpler requirement would logically be subserved by correspond-

ingly simpler sensory coding, as manifested in the invariant responses

observed from auditory neurons.

In summary, we have shown that auditory processing is functional

in larval zebrafish, and that distinct but functionally overlapping catego-

ries of auditory responsive neurons carry out this processing. The over-

all architecture of the system resembles that of mammals, although the

range of frequencies detected, the frequency selectivity of individual

neurons, and the topographic representation of frequency are all rudi-

mentary. The major brain regions involved, the apparent flow of infor-

mation among these regions, and the encoding of the stimulus

intensity are all similar to that seen in adult fish and mammals. These

results suggest that the larval zebrafish will be a suitable platform for

exploring the ways in which auditory processing develops and integra-

tes with the processing of other sensory modalities.

ACKNOWLEDGMENTS

We thank Misha Ahrens for the elavl3:H2B:GCaMP6f transgenic line.

Support was provided by an NHMRC Project Grant (APP1066887),

ARC Future Fellowship (FT110100887), a Simons Foundation

Explorer Award (336331), and two ARC Discovery Project Grants

(DP140102036 & DP110103612) to E.K.S.; an EMBO Long-term

Fellowship to G.V; and an Australian Postgraduate Award to L.A.H.

Imaging work was performed in the Queensland Brain Institute’s

Advanced Microscopy Facility and generously supported by ARC

LIEF LE130100078.

CONFLICT OF INTEREST

None declared.

REFERENCES

Akerboom, J., Chen, T. W., Wardill, T. J., Tian, L., Marvin, J. S., Mutlu, S.,. . . Looger, L. L. (2012). Optimization of a GCaMP calcium indicatorfor neural activity imaging. Journal of Neuroscience, 32(40), 13819–13840. doi:10.1523/JNEUROSCI.2601-12.2012

Beck, J. C., Gilland, E., Tank, D. W., & Baker, R. (2004). Quantifying the

ontogeny of optokinetic and vestibuloocular behaviors in zebrafish,

medaka, and goldfish. Journal of Neurophysiology, 92(6), 3546–3561.doi:10.1152/jn.00311.2004

Bever, M. M., & Fekete, D. M. (2002). Atlas of the developing inner ear

in zebrafish. Developmental Dynamics, 223(4), 536–543. doi:10.1002/dvdy.10062

Bhandiwad, A. A., Zeddies, D. G., Raible, D. W., Rubel, E. W., & Sisneros,

J. A. (2013). Auditory sensitivity of larval zebrafish (Danio rerio) meas-

ured using a behavioral prepulse inhibition assay. Journal of Experimen-

tal Biology, 216(Pt 18), 3504–3513. doi:10.1242/jeb.087635

Bianco, I. H., & Engert, F. (2015). Visuomotor transformations underlying

hunting behavior in zebrafish. Current Biology, 25(7), 831–846. doi:10.1016/j.cub.2015.01.042

Bianco, I. H., Kampff, A. R., & Engert, F. (2011). Prey capture behavior

evoked by simple visual stimuli in larval zebrafish. Frontiers in Systems

Neuroscience, 5, 101. doi:10.3389/fnsys.2011.00101

Bohm, U. L., Prendergast, A., Djenoune, L., Nunes Figueiredo, S., Gomez,

J., Stokes, C., . . . Wyart, C. (2016). CSF-contacting neurons regulate

locomotion by relaying mechanical stimuli to spinal circuits. Nature

Communications, 7, 10866. doi:10.1038/ncomms10866

Bourk, T. R., Mielcarz, J. P., & Norris, B. E. (1981). Tonotopic organization

of the anteroventral cochlear nucleus of the cat. Hearing Research, 4

(3–4), 215–241. doi: Doi 10.1016/0378–5955(81)90008–3

Cadotte, A. J., DeMarse, T. B., He, P., & Ding, M. (2008). Causal meas-

ures of structure and plasticity in simulated and living neural net-

works. PLoS One, 3(10), e3355. doi:10.1371/journal.pone.0003355

Cervi, A. L., Poling, K. R., & Higgs, D. M. (2012). Behavioral measure of

frequency detection and discrimination in the zebrafish, Danio rerio.

Zebrafish, 9(1), 1–7. doi:10.1089/zeb.2011.0720

Chen, T. W., Wardill, T. J., Sun, Y., Pulver, S. R., Renninger, S. L., Baohan, A.,

. . . Kim, D. S. (2013). Ultrasensitive fluorescent proteins for imaging neu-

ronal activity. Nature, 499(7458), 295–300. doi:10.1038/nature12354

Del Bene, F., Wyart, C., Robles, E., Tran, A., Looger, L., Scott, E. K., . . .

Baier, H. (2010). Filtering of visual information in the tectum by an

identified neural circuit. Science, 330(6004), 669–673. doi:10.1126/science.1192949

Demanez, J. P., & Demanez, L. (2003). Anatomophysiology of the central

auditory nervous system: basic concepts. Acta Oto-Rhino-Laryngologica

Belgica, 57(4), 227–236.

Douglass, A. D., Kraves, S., Deisseroth, K., Schier, A. F., & Engert, F.

(2008). Escape behavior elicited by single, channelrhodopsin-2-

evoked spikes in zebrafish somatosensory neurons. Current Biology,

18(15), 1133–1137. doi:10.1016/j.cub.2008.06.077

Dubbs, A., Guevara, J., & Yuste, R. (2016). moco: Fast motion correction

for calcium imaging. Frontiers in Neuroinformatics, 10, 6. doi:10.3389/

fninf.2016.00006

Dunn, T. W., Gebhardt, C., Naumann, E. A., Riegler, C., Ahrens, M. B.,

Engert, F., & Del Bene, F. (2016). Neural circuits underlying visually

evoked escapes in larval zebrafish. Neuron, 89(3), 613–628. doi:

10.1016/j.neuron.2015.12.021

Easter, S. S., Jr., & Nicola, G. N. (1997). The development of eye movements in

the zebrafish (Danio rerio). Developmental Psychobiology, 31(4), 267–276.

Echteler, S. M. (1984). Connections of the auditory midbrain in a teleost

fish, Cyprinus carpio. Journal of Comparative Neurology, 230(4), 536–551. doi:10.1002/cne.902300405

Echteler, S. M. (1985a). Organization of central auditory pathways in a

teleost fish, Cyprinus-Carpio. Journal of Comparative Physiology A-

Sensory Neural and Behavioral Physiology, 156(2), 267–280. doi:

10.1007/Bf00610868

Echteler, S. M. (1985b). Tonotopic organization in the midbrain of a tele-

ost fish. Brain Research, 338(2), 387–391. doi:Doi 10.1016/0006–8993(85)90175–1

Edelstein, A., Amodaj, N., Hoover, K., Vale, R., & Stuurman, N. (2010).

Computer control of microscopes using microManager. Current Proto-

cols in Molecular Biology, Chapter 14, Unit14 20. doi:10.1002/

0471142727.mb1420s92

Farago, A. F., Awatramani, R. B., & Dymecki, S. M. (2006). Assembly of

the brainstem cochlear nuclear complex is revealed by intersectional

and subtractive genetic fate maps. Neuron, 50(2), 205–218. doi:

10.1016/j.neuron.2006.03.014

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 11

Fay, R. R., & Edds-Walton, P. L. (2008). Structures and functions of the

auditory nervous system of fishes. In J. F. Webb, R. R. Fay, & A. N.

Popper (Eds.), Fish Bioacoustics: With 81 Illustrations (pp. 49–97).New York, NY: Springer.

Fidelin, K., Djenoune, L., Stokes, C., Prendergast, A., Gomez, J., Baradel,

A., . . . Wyart, C. (2015). State-dependent modulation of locomotion

by GABAergic spinal sensory neurons. Current Biology, 25(23), 3035–3047. doi:10.1016/j.cub.2015.09.070

Filosa, A., Barker, A. J., Dal Maschio, M., & Baier, H. (2016). Feeding

state modulates behavioral choice and processing of prey stimuli in

the Zebrafish Tectum. Neuron, 90(3), 596–608. doi:10.1016/j.

neuron.2016.03.014

Frisina, R. D. (2001). Subcortical neural coding mechanisms for auditory

temporal processing. Hearing Research, 158(1–2), 1–27. doi:10.1016/S0378-5955(01)00296-9

Fujiyama, T., Yamada, M., Terao, M., Terashima, T., Hioki, H., Inoue, Y.

U., . . . Hoshino, M. (2009). Inhibitory and excitatory subtypes of

cochlear nucleus neurons are defined by distinct bHLH transcription

factors, Ptf1a and Atoh1. Development, 136(12), 2049–2058. doi:

10.1242/dev.033480

Gahtan, E., Tanger, P., & Baier, H. (2005). Visual prey capture in larval

zebrafish is controlled by identified reticulospinal neurons down-

stream of the tectum. Journal of Neuroscience, 25(40), 9294–9303.doi:10.1523/JNEUROSCI.2678-05.2005

Grande, T., & Young, B. (2004). The ontogeny and homology of the

Weberian apparatus in the zebrafish Danio rerio (Ostariophysi: Cypri-

niformes). Zoological Journal of the Linnean Society, 140(2), 241–254.doi: 10.1111/j.1096-3642.2003.00097.x

Granger, C. W. J. (1969). Investigating causal relations by econometric

models and cross-spectral methods. Econometrica, 37(3), 424–438.

Grothe, B., Pecka, M., & McAlpine, D. (2010). Mechanisms of sound

localization in mammals. Physiological Reviews, 90(3), 983–1012. doi:10.1152/physrev.00026.2009

Harris, J. A., Cheng, A. G., Cunningham, L. L., MacDonald, G., Raible, D.

W., & Rubel, E. W. (2003). Neomycin-induced hair cell death and

rapid regeneration in the lateral line of zebrafish (Danio rerio). Journal

of the Association for Research in Otolaryngology, 4(2), 219–234. doi:10.1007/s10162-002-3022-x

Higgs, D. M., Rollo, A. K., Souza, M. J., & Popper, A. N. (2003). Develop-

ment of form and function in peripheral auditory structures of the

zebrafish (Danio rerio). Journal of the Acoustical Society of America,

113(2), 1145–1154.

Higgs, D. M., Souza, M. J., Wilkins, H. R., Presson, J. C., & Popper, A. N.

(2002). Age- and size-related changes in the inner ear and hearing

ability of the adult zebrafish (Danio rerio). Journal of the Association

for Research in Otolaryngology, 3(2), 174–184.

Jia, H., Rochefort, N. L., Chen, X., & Konnerth, A. (2011). In vivo two-

photon imaging of sensory-evoked dendritic calcium signals in cortical

neurons. Nature Protocols, 6(1), 28–35. doi:10.1038/nprot.2010.169

Johansen, S. (1991). Estimation and hypothesis-testing of cointegration

vectors in Gaussian vector autoregressive models. Econometrica, 59

(6), 1551–1580. doi: 10.2307/2938278

Kalmijn, A. J. (1988). Hydrodynamic and acoustic field detection. In J.

Atema, R. R. Fay, A. N. Popper, & W. N. Tavolga (Eds.), Sensory biol-

ogy of aquatic animals (pp. 83–130). New York, NY: Springer.

Kandler, K., Clause, A., & Noh, J. (2009). Tonotopic reorganization of

developing auditory brainstem circuits. Nature Neuroscience, 12(6),

711–717. doi:10.1038/nn.2332

Kinkhabwala, A., Riley, M., Koyama, M., Monen, J., Satou, C., Kimura, Y.,

. . . Fetcho, J. (2011). A structural and functional ground plan for

neurons in the hindbrain of zebrafish. Proceedings of the National

Academy of Sciences of the United States of America, 108(3), 1164–1169. doi:10.1073/pnas.1012185108

Kirsch, J. A., Hofmann, M. H., Mogdans, J., & Bleckmann, H. (2002).

Responses of diencephalic neurons to sensory stimulation in the

goldfish, Carassius auratus. Brain Research Bulletin, 57(3–4), 419–421.doi:10.1016/S0361-9230(01)00703-1

Kwiatkowski, D., Phillips, P. C. B., Schmidt, P., & Shin, Y. C. (1992). Testing

the null hypothesis of stationarity against the alternative of a unit-root -

How sure are we that economic time-series have a unit-root. Journal of

Econometrics, 54(1–3), 159–178. doi: 10.1016/0304-4076(92)90104-Y

Lai, J. S., Lo, S. J., Dickson, B. J., & Chiang, A. S. (2012). Auditory circuitin the Drosophila brain. Proceedings of the National Academy of Scien-ces of the United States of America, 109(7), 2607–2612. doi:10.1073/pnas.1117307109

Legland, D., Arganda-Carreras, I., & Andrey, P. (2016). MorphoLibJ: Inte-

grated library and plugins for mathematical morphology with ImageJ.

Bioinformatics, 32(22), 3532–3534. doi:10.1093/bioinformatics/btw413

Levi, R., Akanyeti, O., Ballo, A., & Liao, J. C. (2015). Frequency response

properties of primary afferent neurons in the posterior lateral line

system of larval zebrafish. Journal of Neurophysiology, 113(2), 657–668. doi:10.1152/jn.00414.2014

Liao, J. C., & Haehnel, M. (2012). Physiology of afferent neurons in larvalzebrafish provides a functional framework for lateral line somatotopy. Jour-nal of Neurophysiology, 107(10), 2615–2623. doi:10.1152/jn.01108.2011

Lippe, W., & Rubel, E. W. (1985). Ontogeny of tonotopic organization ofbrain stem auditory nuclei in the chicken: Implications for develop-ment of the place principle. Journal of Comparative Neurology, 237(2),273–289. doi:10.1002/cne.902370211

Lippe, W. R. (1987). Shift of tonotopic organization in brain stem audi-

tory nuclei of the chicken during late embryonic development. Hear

Research, 25(2–3), 205–208. doi: 10.1016/0378-5955(87)90092-X

Lombarte, A., Yan, H. Y., Popper, A. N., Chang, J. S., & Platt, C. (1993).

Damage and regeneration of hair cell ciliary bundles in a fish ear fol-

lowing treatment with gentamicin. Hear Research, 64(2), 166–174.

Lu, Z., & DeSmidt, A. A. (2013). Early development of hearing in zebra-

fish. Journal of the Association for Research in Otolaryngology, 14(4),

509–521. doi:10.1007/s10162-013-0386-z

McCormick, C. A., & Hernandez, D. V. (1996). Connections of octaval and

lateral line nuclei of the medulla in the goldfish, including the cytoarchi-

tecture of the secondary octaval population in goldfish and catfish. Brain,

Behavior and Evolution, 47(3), 113–137. doi:Doi 10.1159/000113232

Meredith, G. E., & Butler, A. B. (1983). Organization of eighth nerve

afferent projections from individual endorgans of the inner ear in the

teleost, Astronotus ocellatus. Journal of Comparative Neurology, 220

(1), 44–62. doi:10.1002/cne.902200106

Micheyl, C., Schrater, P. R., & Oxenham, A. J. (2013). Auditory frequency

and intensity discrimination explained using a cortical population rate

code. PLOS Computational Biology, 9(11), e1003336. doi:10.1371/

journal.pcbi.1003336

Mirjany, M., Preuss, T., & Faber, D. S. (2011). Role of the lateral line

mechanosensory system in directionality of goldfish auditory evoked

escape response. Journal of Experimental Biology, 214(Pt 20), 3358–3367. doi:10.1242/jeb.052894

Moore, B. C. (2003). Coding of sounds in the auditory system and its

relevance to signal processing and coding in cochlear implants. Otol-

ogy & Neurotology, 24(2), 243–254.

Mu, Y., Li, X. Q., Zhang, B., & Du, J. L. (2012). Visual input modulates

audiomotor function via hypothalamic dopaminergic neurons through

a cooperative mechanism. Neuron, 75(4), 688–699. doi:10.1016/j.

neuron.2012.05.035

12 | The Journal ofComparative Neurology

VANWALLEGHEM ET AL.

Mueller, T. (2012). What is the Thalamus in Zebrafish? Frontiers in Neuro-

science, 6, 64. doi:10.3389/fnins.2012.00064

Mueller, T., & Wullimann, M. F. (2016). Chapter 2 - Atlas of cellular

markers in Zebrafish Neurogenesis: Atlas. In Atlas of Early Zebrafish

Brain Development (2nd edn.) (pp. 27–157). San Diego: Elsevier.

Muniak, M. A., & Ryugo, D. K. (2014). Tonotopic organization of vertical

cells in the dorsal cochlear nucleus of the CBA/J mouse. Journal of

Comparative Neurology, 522(4), 937–949. doi:10.1002/cne.23454

Northcutt, R. G. (1981). Audition and the central nervous system of fishes.

In W. N. Tavolga, A. N. Popper, & R. R. Fay (Eds.), Hearing and Sound

Communication in Fishes (pp. 331–355). New York, NY: Springer.

Northcutt, R. G. (2006). Connections of the lateral and medial divisions

of the goldfish telencephalic pallium. Journal of Comparative Neurol-

ogy, 494(6), 903–943. doi:10.1002/cne.20853

Orger, M. B., & Baier, H. (2005). Channeling of red and green cone

inputs to the zebrafish optomotor response. Visual Neuroscience, 22

(3), 275–281. doi:10.1017/S0952523805223039

Orger, M. B., Smear, M. C., Anstis, S. M., & Baier, H. (2000). Perception

of Fourier and non-Fourier motion by larval zebrafish. Nature Neuro-

science, 3(11), 1128–1133. doi:10.1038/80649

Popper, A. N., & Fay, R. R. (1993). Sound detection and processing by

fish: critical review and major research questions. Brain, Behavior and

Evolution, 41(1), 14–38. doi: 10.1159/000113821

Popper, A. N., & Fay, R. R. (1999). The auditory periphery in fishes. In R.

R. Fay & A. N. Popper (Eds.), Comparative hearing: Fish and amphib-

ians (pp. 43–100). New York, NY: Springer.

Preibisch, S., Saalfeld, S., & Tomancak, P. (2009). Globally optimal stitch-

ing of tiled 3D microscopic image acquisitions. Bioinformatics, 25(11),

1463–1465. doi:10.1093/bioinformatics/btp184

Raible, D. W., & Kruse, G. J. (2000). Organization of the lateral line system in

embryonic zebrafish. Journal of Comparative Neurology, 421(2), 189–198.

Randlett, O., Wee, C. L., Naumann, E. A., Nnaemeka, O., Schoppik, D.,

Fitzgerald, J. E., . . . Schier, A. F. (2015). Whole-brain activity mapping

onto a zebrafish brain atlas. Nat Methods, 12(11), 1039–1046. doi:10.1038/nmeth.3581

Riley, B. B., & Moorman, S. J. (2000). Development of utricular otoliths,

but not saccular otoliths, is necessary for vestibular function and sur-

vival in zebrafish. Journal of Neurobiology, 43(4), 329–337.

Roeser, T., & Baier, H. (2003). Visuomotor behaviors in larval zebrafish

after GFP-guided laser ablation of the optic tectum. Journal of Neuro-

science, 23(9), 3726–3734.

Rohmann, K. N., Tripp, J. A., Genova, R. M., & Bass, A. H. (2014). Manip-

ulation of BK channel expression is sufficient to alter auditory hair

cell thresholds in larval zebrafish. Journal of Experimental Biology, 217

(Pt 14), 2531–2539. doi:10.1242/jeb.103093

Sachs, M. B., & Abbas, P. J. (1974). Rate versus level functions for

auditory-nerve fibers in cats: Tone-burst stimuli. Journal of the Acous-

tical Society of America, 56(6), 1835–1847.

Said, S. E., & Dickey, D. A. (1984). Testing for unit roots in

autoregressive-moving average models of unknown order. Biometrika,

71(3), 599–607.

Seth, A. K., Barrett, A. B., & Barnett, L. (2015). Granger causality analysis

in neuroscience and neuroimaging. Journal of Neuroscience, 35(8),

3293–3297. doi:10.1523/JNEUROSCI.4399-14.2015

Simmons, A. M., & Flores, V. (2012). Particle motion is broadly repre-

sented in the vestibular medulla of the bullfrog across larval develop-

ment. Journal of Comparative Physiology A, 198(4), 253–266. doi:

10.1007/s00359-011-0705-y

Tanimoto, M., Ota, Y., Horikawa, K., & Oda, Y. (2009). Auditory input to

CNS is acquired coincidentally with development of inner ear after

formation of functional afferent pathway in zebrafish. Journal of Neu-

roscience, 29(9), 2762–2767. doi:10.1523/JNEUROSCI.5530-08.2009

Temizer, I., Donovan, J. C., Baier, H., & Semmelhack, J. L. (2015). a visual

pathway for looming-evoked escape in larval zebrafish. Current Biol-

ogy, 25(14), 1823–1834. doi:10.1016/j.cub.2015.06.002

Thevenaz, P., Ruttimann, U. E., & Unser, M. (1998). A pyramid approach

to subpixel registration based on intensity. IEEE Transactions on Image

Processing, 7(1), 27–41. doi:10.1109/83.650848

Thompson, A. W., & Scott, E. K. (2016). Characterisation of sensitivity

and orientation tuning for visually responsive ensembles in the zebra-

fish tectum. Scientific Reports, 6, 34887. doi:10.1038/srep34887

Thompson, A. W., Vanwalleghem, G. C., Heap, L. A., & Scott, E. K.

(2016). Functional profiles of visual-, auditory-, and water flow-

responsive neurons in the zebrafish tectum. Current Biology, 26(6),

743–754. doi:10.1016/j.cub.2016.01.041

Toda, H. Y., & Yamamoto, T. (1995). Statistical-inference in vector autor-

egressions with possibly integrated processes. Journal of Econome-

trics, 66(1–2), 225–250. doi:Doi 10.1016/0304–4076(94)01616–8

Vladimirov, N., Mu, Y., Kawashima, T., Bennett, D. V., Yang, C. T., Looger, L. L.,

. . . Ahrens, M. B. (2014). Light-sheet functional imaging in fictively behav-

ing zebrafish. Nature Methods, 11(9), 883–884. doi:10.1038/nmeth.3040

Wang, J., Song, Q., Yu, D., Yang, G., Xia, L., Su, K., . . . Yin, S. (2015).

Ontogenetic development of the auditory sensory organ in zebrafish

(Danio rerio): Changes in hearing sensitivity and related morphology.

Scientific Reports, 5, 15943. doi:10.1038/srep15943

Webster, D. B. (1992). An overview of mammalian auditory pathways

with an emphasis on humans. In D. B. Webster, A. N. Popper, & R. R.

Fay (Eds.), The mammalian auditory pathway: Neuroanatomy (pp. 1–22). New York, NY: Springer.

Wen, L., Wei, W., Gu, W., Huang, P., Ren, X., Zhang, Z., . . . Zhang, B.

(2008). Visualization of monoaminergic neurons and neurotoxicity of

MPTP in live transgenic zebrafish. Developmental Biology, 314(1), 84–92. doi:10.1016/j.ydbio.2007.11.012

Westerfield, M. (2000). The zebrafish book. A guide for the laboratory use of

zebrafish (Danio rerio). (4th ed.). Eugene, OR: University of Oregon Press.

Wullimann, M. F., Mueller, T., Distel, M., Babaryka, A., Grothe, B., & Kos-

ter, R. W. (2011). The long adventurous journey of rhombic lip cells

in jawed vertebrates: A comparative developmental analysis. Frontiers

in Neuroanatomy, 5, 27. doi:10.3389/fnana.2011.00027

Yao, Q., DeSmidt, A. A., Tekin, M., Liu, X., & Lu, Z. (2016). Hearing

assessment in zebrafish during the first week postfertilization. Zebra-

fish, 13(2), 79–86. doi:10.1089/zeb.2015.1166

SUPPORTING INFORMATION

Additional Supporting Information may be found online in the sup-

porting information tab for this article.

TABLE S1. Acceleration measurement of tones.

How to cite this article: Vanwalleghem G, Heap LA, Scott EK. A

profile of auditory-responsive neurons in the larval zebrafish

brain. J Comp Neurol. 2017;00:1–13. https://doi.org/10.1002/

cne.24258

VANWALLEGHEM ET AL. The Journal ofComparative Neurology

| 13