Delayed Otolith Development Does Not Impair Vestibular Circuit Formation in Zebrafish

Abstract

What is the role of normally patterned sensory signaling in development of vestibular circuits? For technical reasons, including the difficulty in depriving animals of vestibular inputs, this has been a challenging question to address. Here we take advantage of a vestibular-deficient zebrafish mutant, rock solo ^AN66 , in order to examine whether normal sensory input is required for formation of vestibular-driven postural circuitry. We show that the rock solo ^AN66 mutant is a splice site mutation in the secreted glycoprotein otogelin (otog), which we confirm through both whole genome sequencing and complementation with an otog early termination mutant. Using confocal microscopy, we find that elements of postural circuits are anatomically normal in rock solo ^AN66 mutants, including hair cells, vestibular ganglion neurons, and vestibulospinal neurons. Surprisingly, the balance and postural deficits that are readily apparent in younger larvae disappear around 2 weeks of age. We demonstrate that this behavioral recovery follows the delayed development of the anterior (utricular) otolith, which appears around 14 days post-fertilization (dpf), compared to 1 dpf in WT. These findings indicate that utricular signaling is not required for normal structural development of the inner ear and vestibular nucleus neurons. Furthermore, despite the otolith's developmental delay until well after postural behaviors normally appear, downstream circuits can drive righting reflexes within ∼1-2 days of its arrival, indicating that vestibular circuit wiring is not impaired by a delay in patterned activity. The functional recovery of postural behaviors may shed light on why humans with mutations in otog exhibit only subclinical vestibular deficits.

Keywords: critical period; development; otolith; posture; vestibulospinal; zebrafish.

FIG.1

A mutation in a splice site of otogelin underlies rock solo ^AN66 phenotype. a Top: exon structure of otog and location of the splice site mutation in rock solo ^AN66 (first asterisk) and the early termination mutation in the otog ^sa10228 line (second asterisk). Bottom: electropherograms of otog sequencing in the rock solo ^AN66 (left) and otog ^sa10228 (right) lines at the loci identified. Data are from pooled larvae. b Transmitted light images of the ear, viewed from the lateral aspect, in 5–7 dpf larvae. The anterior (utricular) otolith (dashed outline) is absent in both the rock solo ^AN66 homozygotes as well as in offspring from the complementation cross between rock solo ^AN66 and otog ^sa10228. Rostral is to the left and dorsal is up.

FIG. 2

No change in the number of hair cells in the utricular macula in otog ^c.1522+2T>A animals. a Confocal image of the anterior/utricular macula in WT/heterozygous siblings. Hair cells are stained with the HCS-1 antibody to otoferlin. Image is the average of 15 z-planes totaling 11 μm thick, encompassing the whole macula. Anterior is to the left, dorsal to the top of this and all subsequent images. b As in a, for a otog ^c.1522+2T>A−/− animal. Image is the average of seven z-planes totaling 5 μm. The small cluster of hair cells in upper right of the image is from a lateral line neuromast. Cells appear slightly more oval than in WT because of a modest difference in mounting angle. c Counts of the number of utricular hair cells in WT/heterozygous and otog ^c.1522+2T>A animals. Each dot represents data from a different animal.

FIG. 3

No change in vestibular ganglion neurons in otog ^c.1522+2T>A animals. a Confocal image of the vestibular ganglion as visualized in the Tg (isl2b:GFP) line in WT/heterozygous siblings. Blue, DAPI counter-stain. Image is the average of 11 z-planes totaling 8 μm thick. b As in (a), for a otog ^c.1522+2T>A−/− animal. Image is the average of ten z-planes totaling 7 μm. c Counts of the number of vestibular ganglion neurons in WT/heterozygous and otog ^c.1522+2T>A animals. Each dot represents data from a different animal. d–e Example images of putative synaptic contacts from vestibular ganglion neurons onto utricular hair cells in Tg (isl2b:GFP) WT/heterozygous (d) and otog ^c.1522+2T>A (e) animals. Hair cells (magenta) are labeled with the HCS-1 otoferlin antibody as in Fig. 2. Close punctate apposition between ganglion processes and hair cells can be seen in the merged pictures (right), consistent with the known Type II (bouton-like) anatomy (Eatock and Songer, 2011).

FIG. 4

No change in the number of vestibulospinal neurons in otog ^c.1522+2T>A animals. a Confocal image of retrogradely labeled vestibulospinal neurons in the hindbrain of a WT/heterozygous animal, viewed from the lateral aspect. Image is depth-coded with color reflecting the distance from the lateral-most image. M, Mauthner lateral dendrite, which is seen in cross-section. 1–6, vestibulospinal neurons. b As in (a), for a otog ^c.1522+2T>A animal. c Counts of the number of vestibulospinal neurons labeled in WT/heterozygous and otog ^c.1522+2T>A animals. Each dot represents data from a different animal.

FIG. 5

Delayed otolith development triggers behavioral recovery. a Transmitted light images of the otoliths of the inner ear in three animals. In a WT animal (top), the utricular (U), saccular (S), and lagenar (L) otoliths are well-formed by 13 dpf. A typical otog ^c.1522+2T>A animal exhibits only the S and L otoliths at 13 dpf (middle). Around 14 dpf, many otog ^c.1522+2T>A animals spontaneously develop a utricular otolith, which is initially smaller than that of WT (bottom). The bottom image is stitched from two separate sets of z-planes due to the small size of the nascent utricular otolith. b Examples of abnormal and normal swimming posture in one otog ^c.1522+2T>A larva. This animal first developed utricular otoliths at 14 dpf, but on that day still swam abnormally, with dorsal-up posture on 0/5 swim bouts (top); here it is seen from above lying left side up and then reversing to orient right flank up during a swim. The plastic pipette used to deliver a touch stimulus can be seen contacting the head at the beginning of swim. At 15 dpf (bottom) this same animal began and maintained its swim with dorsal-up posture in 5/5 swim bouts. A 45 ° mirror can be seen at the top of these images, confirming that this animal is dorsal up. Images represent subsets of frames from high-speed videography under infrared illumination and are 2 ms exposures, 14 ms apart. c Summary of postural behaviors before and after the delayed arrival of 1 or 2 utricular otoliths in 14 otog ^c.1522+2T>A animals. Day 0 indicates the first day on which at least one utricular otolith was observed in approximately normal position, regardless of size (actual age at day 0, 13–17 dpf). Animals that maintained dorsal-up orientation throughout swim bouts in the dark were considered to have normal posture; animals that rolled sideways, upside-down, or pitched vertically during 50 % or more of swim bouts were considered to have abnormal posture. Each row represents observations from a different animal. Examples of swimming behavior are shown in b for the animal highlighted with red circles. d Postural behaviors before and after the development of lagenar otoliths in four animals. Day 0 indicates the first day on which at least one normally positioned lagenar otolith develops (actual age, 13–15 dpf). In these animals, which did not develop utricular otoliths over the time course represented here, posture remained abnormal.