The lhfpl5 Ohnologs lhfpl5a and lhfpl 5b Are Required for Mechanotransduction in Distinct Populations of Sensory Hair Cells in Zebrafish

Abstract

Hair cells sense and transmit auditory, vestibular, and hydrodynamic information by converting mechanical stimuli into electrical signals. This process of mechano-electrical transduction (MET) requires a mechanically gated channel localized in the apical stereocilia of hair cells. In mice, lipoma HMGIC fusion partner-like 5 (LHFPL5) acts as an auxiliary subunit of the MET channel whose primary role is to correctly localize PCDH15 and TMC1 to the mechanotransduction complex. Zebrafish have two lhfpl5 genes (lhfpl5a and lhfpl5b), but their individual contributions to MET channel assembly and function have not been analyzed. Here we show that the zebrafish lhfpl5 genes are expressed in discrete populations of hair cells: lhfpl5a expression is restricted to auditory and vestibular hair cells in the inner ear, while lhfpl5b expression is specific to hair cells of the lateral line organ. Consequently, lhfpl5a mutants exhibit defects in auditory and vestibular function, while disruption of lhfpl5b affects hair cells only in the lateral line neuromasts. In contrast to previous reports in mice, localization of Tmc1 does not depend upon Lhfpl5 function in either the inner ear or lateral line organ. In both lhfpl5a and lhfpl5b mutants, GFP-tagged Tmc1 and Tmc2b proteins still localize to the stereocilia of hair cells. Using a stably integrated GFP-Lhfpl5a transgene, we show that the tip link cadherins Pcdh15a and Cdh23, along with the Myo7aa motor protein, are required for correct Lhfpl5a localization at the tips of stereocilia. Our work corroborates the evolutionarily conserved co-dependence between Lhfpl5 and Pcdh15, but also reveals novel requirements for Cdh23 and Myo7aa to correctly localize Lhfpl5a. In addition, our data suggest that targeting of Tmc1 and Tmc2b proteins to stereocilia in zebrafish hair cells occurs independently of Lhfpl5 proteins.

Keywords: LHFPL5; PCDH15; TMC1; deafness; hair cell; lateral line; mechanotransduction; zebrafish.

FIGURE 1

Duplicated lhfpl5 genes in teleost fish. (A) Phylogenic tree of representative teleost and vertebrate Lhfpl5 protein sequences. See Supplementary Table S1 for species’ names and protein accession numbers. (B) Sequence alignment of Lhfpl5 proteins from zebrafish (Danio rerio, Dr), chicken (Gallus gallus, Gg), mouse (Mus musculus, Mm), and humans (Homo sapiens, Hs). Regions of the proteins coded for by exons 1 and 3 are shaded gray. Locations of the transmembrane (TM) helices 1–4 are shown in the linear protein structure diagram. The lhfpl5a^tm290d and lhfpl5b^vo35 mutations are indicated by the red triangle and red square, respectively.

FIGURE 2

Zebrafish lhfpl5a and lhfpl5b genes are expressed in distinct populations of sensory hair cells. Whole mount mRNA in situ hybridization for lhfpl5a (A,C,E,G,I) and lhfpl5b (B,D,F,H,J) at 1, 2, and 5 days post-fertilization (dpf). (G–J) Are details from 5 dpf larvae. NMs, neuromasts; HC, hair cell; SC, support cell. Scale bars: (A,B) = 50 μm; (C–F) = 100 μm; (G–J) = 25 μm.

FIGURE 3

lhfpl5a is required for auditory and vestibular hair cell function. (A) Boxplot showing the first peak amplitude (μV) of microphonic recordings from the inner ear of 3 dpf wild type, lhfpl5a^tm290d, and lhfpl5b^vo35 larvae. The boxes cover the inter-quartile range (IQR), and the whiskers represent the minimum and maximum datapoints within 1.5 times the IQR. Values from individual larvae are indicated by the black dots and outliers indicated by a diamond. Asterisks indicate p < 0.01 (^∗∗) by Welch’s t-test. ns, not significant. (B) Boxplot of the acoustic startle response in 6 dpf wild type and lhfpl5a^tm290d mutants, with or without the GFP-lhfpl5a vo23Tg transgene. Values from individual larvae are indicated by the black dots and outliers indicated by a diamond. Asterisks indicate p < 0.001 (^∗∗∗) by Welch’s t-test. ns, not significant. (C–F’) FM 4–64 dye labeling assay for MET channel activity of inner ear hair cells from 6 dpf larvae [C – Wild type (+/+ or±), D – lhfpl5a^tm290d mutants, E – Wild type GFP-lhfpl5a vo23Tg, F – GFP-lhfpl5a vo23Tg; lhfpl5a^tm290]. (E’,F’) Are images of GFP-Lhfpl5a protein in the bundle of the hair cells shown above in (E,F). n = 2, 3, 12, and 4 for the genotypes in (C–F), respectively. Scale bars = 5 μm and apply to all images.

FIGURE 4

lhfpl5b is required for lateral line hair cell function. (A–C) Representative images of 5 dpf zebrafish larvae (A – Wild type; B – lhfpl5a^tm290d; C – lhfpl5b^vo35) labeled with the MET channel-permeant dye FM 1–43. (D–G) Representative images of individual neuromasts from 2 and 5 dpf wild type and lhfpl5b^vo35 larvae labeled with FM 1–43. Dashed lines outline the cluster of hair cells in each neuromast. (H) Quantification of normalized FM 1–43 fluorescence intensity per hair cell of 2 and 5 dpf neuromasts (n = 10 WT, 14 lhfpl5b^vo35 NMs at 2 dpf; n = 6 WT, 13 lhfpl5b^vo35 NMs each genotype at 5 dpf). The box plots cover the inter-quartile range (IQR), and the whiskers represent the minimum and maximum datapoints within 1.5 times the IQR. Asterisks indicate p < 0.001 (^∗∗∗) by Welch’s t-test. (I,J) Rescue of FM dye labeling in lhfpl5b^vo35 mutants (n = 7) by the GFP-lhfpl5a (vo23Tg) transgene. The GFP-Lhfpl5a bundle and FM 4–64 images are from the same NM for each genotype. (K) Quantification of hair cell number in L1, MI1, and O2 neuromasts from 5 dpf lhfpl5b^vo35 mutants (n = 11) and wild-type siblings (n = 11). The box plots are the same as in (H). Asterisks indicate p < 0.001 (^∗∗∗) by Welch’s t-test. Scale bars = 5 μm, applies to (D–G,I,J); 2 μm in (I,J) insets.

FIGURE 5

Tmc proteins do not require Lhfpl5a or Lhfpl5b for localization to the stereocilia of zebrafish hair cells. (A–D’) Representative images of Tmc1-GFP (vo27Tg) or Tmc2b-GFP (vo28Tg) in the lateral cristae of wild type (A,A’,C,C’) and lhfpl5a^tm290d (B,B’,D,D’) larvae. The GFP-only channel is shown in (A–D) and overlaid with a light image of the bundles in (A’–D’). (E–F’) Representative images of Tmc2b-GFP (vo28Tg) in the neuromasts of wild type (E,E’) and lhfpl5b^vo35 (F,F’) larvae. The GFP-only channel is shown in (E,F) and overlaid with a light image of the bundles in (E’,F’). Scale bars = 3 μm in all panels.

FIGURE 6

Lhfpl5a requires MET complex proteins Pcdh15a, Cdh23, and Myo7a for normal localization in the stereocilia of zebrafish hair cells. (A,B) Immunostain of Pcdh15a (magenta) in the lateral cristae of wild type and lhfpl5a^tm290d mutants at 3 dpf. Phalloidin-stained actin of the hair bundle is shown in green. Arrows indicate areas of Pcdh15a accumulation. (C–F’) Representative images of GFP-Lhfpl5a (vo23Tg) in the lateral cristae hair bundles of wild type (C,C’) and pcdh15a^psi7 (D,D’), cdh23^nl9 (E,E’), and myo7aa^ty220 (F,F’) mutants. White arrows indicate GFP signal in the presumptive kinocilial linkages, yellow arrow heads indicate GFP signal in the stereocilia or the base of the hair bundle, and brackets indicate GFP signal in the kinocilium. Scale bars = 3 μm in (A,B); 2 μm in (C–F’).