Novel degenerative and developmental defects in a zebrafish model of mucolipidosis type IV

Abstract

Mucolipidosis type IV (MLIV) is a lysosomal storage disease characterized by neurologic and ophthalmologic abnormalities. There is currently no effective treatment. MLIV is caused by mutations in MCOLN1, a lysosomal cation channel from the transient receptor potential (TRP) family. In this study, we used genome editing to knockout the two mcoln1 genes present in Danio rerio (zebrafish). Our model successfully reproduced the retinal and neuromuscular defects observed in MLIV patients, indicating that this model is suitable for studying the disease pathogenesis. Importantly, our model revealed novel insights into the origins and progression of the MLIV pathology, including the contribution of autophagosome accumulation to muscle dystrophy and the role of mcoln1 in embryonic development, hair cell viability and cellular maintenance. The generation of a MLIV model in zebrafish is particularly relevant given the suitability of this organism for large-scale in vivo drug screening, thus providing unprecedented opportunities for therapeutic discovery.

Figure 1

Generation of mcoln1 mutants in zebrafish. (A) Alignment of human (hu) MCOLN1, zebrafish (zf) Mcoln1a and zf Mcoln1b protein sequences using online software T-COFFEE. The underline black asterisks mark the consensus sequences among these proteins. The two red rectangular brackets indicate the di-leucine lysosomal targeting motifs. The blue and red asterisks mark the mismatch start for Mcoln1a and Mcoln1b mutants, respectively. (B) Strategy and method used to create mcoln1a and mcoln1b mutants. ZFN target site for mcoln1a is on exon 4, while CRISPR/Cas9 target site for mcoln1b is on exon 2. F1 indicated the primers used to sequence the corresponding cDNAs and confirm the mutations (C) mcoln1a^−/− and mcoln1b^−/− cDNA sequences showing the introduction of a frameshift that results in a premature stop codon for both proteins. (D) Strategy used to create mcoln1ab^−/− animals. (E) Kaplan-Meier survival curve of a cohort study (50 animals per genotype) showing the survival rate of each genotype at 0, 2, 4, 6 and 8 months. Differences between groups were calculated using the log-rank test.

Figure 2

Skeletal muscle pathology in mcoln1 mutant zebrafish at both embryonic and adult stages. (A) Whole mount immunostaining for actinin, vinculin, and α-Bungarotoxin in WT, mcoln1a^−/−, mcoln1b^−/−, and mcoln1ab^−/− zebrafish embryos at 3 dpf. Scale bar, 20 μm. (B) Co-immunostaining for LC3, actinin and DAPI in single muscle fiber dissected from 8-month-old WT and mcoln1ab^−/− zebrafish. White arrows mark the accumulation of LC3 in the middle of the single muscle fiber. Scale bar, 10 μm. (C) Western blot analysis showing LC3 expression levels in skeletal muscle lysates of 8-month-old WT, mcoln1a^−/−, mcoln1b^−/−, and mcoln1ab^−/− zebrafish. Upper band shows the level of LC3-I and lower nad indicated the level of LC3-II. Actin is used as a loading control. (D) Quantification of LC3-II/LC3-I ratio shown in (C). Data are shown as mean ± SD and represent 4 independent experiments. The data were analyzed using paired t test (*P < 0.05). m1a^−/− stands for mcoln1a^−/−, m1b^−/− stands for mcoln1b^−/−, and m1ab^−/− stands for mcoln1ab^−/−. (E) Co-immunostaining for LC3, actinin and DAPI in single muscle fiber dissected from 12-month-old mcoln1ab^−/− zebrafish. White arrow marks the accumulation of LC3 in a split muscle fiber. Scale bar for upper panel is 10 μm, for lower panel is 5 μm. All the images are representative of at least three independent experiments.

Figure 3

Altered morphology and distribution of microtubule, mitochondria, and acethylcholine receptor in mcoln1 zebrafish adult mutant. (A) Co-immunostaining for Cytochrome C, LC3, and DAPI in single muscle fiber dissected from 12-month-old WT and mcoln1ab^−/− zebrafish. (B) Co-staining for actinin, α-Bungarotoxin, and DAPI in single muscle fiber dissected from 12-month-old WT and mcoln1ab^−/− zebrafish. Arrows mark the acetylcholine receptor distribution in mcoln1ab^−/− zebrafish muscle fiber. Scale bar, 10 μm. (C) Co-staining for LysoTracker and DAPI in live single muscle fiber isolated from 12-month-old WT and mcoln1ab^−/− zebrafish. All the images are representative of at least three independent experiments. Scale bar, 50 μm.

Figure 4

Autophagy disruption in skeletal muscle of mcoln1ab^−/− fish (A–G) Muscle samples of 12-month-old mcoln1ab^−/− zebrafish were analyzed by transmission electron microscopy (TEM). Images show autophagic debris containing electro-dense inclusions, organelles and other cytoplasmic contents in mcoln1ab^−/− skeletal muscle. Arrows in (E) indicate multi-lamellar structures; arrowheads in (F) point to abnormal mitochondria; and the asterisks in (G) denotes autophagosomes. Scale bar for (A–F), 1 μm. Scale bar for (G), 250 nm.

Figure 5

Eye pathology in mcoln1 mutant zebrafish at both embryonic and adult stages. (A) H&E staining on paraffin sections of 3 dpf WT and mcoln1ab^−/− zebrafish embryos showing the eye structure. Scale bar, 20 μm. (B) TUNEL and DAPI co-staining on a 3dpf eye paraffin section of mcoln1ab^−/− zebrafish is merged together with an H&E stained section that is 20 μm apart, the white arrow shows a TUNEL positive apoptotic body-like structure. Scale bar, 20 μm. (C) Auto-fluorescence of 8-month-old zebrafish retina on cryostat sections for WT, mcoln1a^−/−, mcoln1b^−/−, and mcoln1ab^−/−. Insets show a twofold magnification of the indicated region. Scale bar, 5 μm. (D) H&E staining on paraffin sections of 8-month-old WT and mcoln1ab^−/− zebrafish retina showing the structure of different retinal layers. The black asterisk indicates thinning of the ONL in mcoln1ab^−/− zebrafish. The black arrows mark cell loss in the INL. RPE, Retinal Pigment Epithelium; PRL, Photoreceptor Layer; ONL, Outer Nuclear Layer; OPL, Outer Plexiform Layer; INL, Inner Nuclear Layer; IPL, Inner Plexiform Layer; RGL, Retinal Ganglion Layer. Scale bar, 10 μm. All the images are representative of at least three independent experiments.

Figure 6

Apoptosis in mcoln1 mutant embryos at 3 dpf. (A) Whole mount TUNEL staining in zebrafish embryos at 3 dpf for WT, mcoln1a^−/−, mcoln1b^−/−, and mcoln1ab^−/− with the orientation to show the eye staining. (B) Quantification of TUNEL positive signals in the eyes of different genotypes. Data are shown as mean ± SD and represent 6 eyes counted. The data were analyzed using paired t test (***P < 0.001; **P < 0.01). (C) Whole mount TUNEL staining in zebrafish embryos at 3 dpf with different genotypes with the orientation to show the brain staining. (D) Quantification of TUNEL positive signals in the head of different genotypes. Data are shown as mean ± SD and represent 3 heads counted. The data were analyzed using paired t test (***P < 0.001).

Figure 7

Hair cell viability and morphology in mcoln1 mutant larvae at 5 dpf. (A) Whole mount YO-PRO-1 staining in live zebrafish larvae at 5 dpf. The positive signal highlighted by the arrow indicates a typical neuromast. Scale bar, 250 μm. (B) A higher magnification of neuromast staining from (A) to show hair cell morphology. Scale bar, 5 μm. (C) Quantification of hair cell number per lateral line neuromast in YO-PRO-1 stained larvae for both WT and mcoln1ab^−/− at 5 dpf. Data are shown as mean ± SD and represent 30 neuromast from 10 embryos counted per genotype in one representative experiment. Three independent experiments were performed. The data were analyzed using paired t test (***P < 0.001). (D) Co-immunostaining for myo7A/HCS-1 and DAPI for fixed WT and mcoln1ab^−/− larvae at 5 dpf. Scale bar, 5 μm. All the images are representative of at least three independent experiments. (E) Reduced number of hair cells in mcoln1ab^−/− head neuromast was rescued by mcoln1a and mcoln1b RNA co-injection, as well as by injection of either mcoln1a RNA or mcoln1b RNA 4 dpf. Data are shown as mean ± SD and represent 10 neuromasts from 10 different embryos per genotype. The result was analyzed using paired t test (***P < 0.001; **P < 0.01; *P < 0.05; n.s. not significant).

Figure 8

Ultrastructure of hair cell in mcoln1ab^−/− larvae at 5 dpf. Transmission Electron Microscopy (TEM) for one neuromast in both WT (A) and mcoln1ab^−/− (B) larvae at 5 dpf. m stands for mantle cell, h stands for hair cell, and s stands for support cell. White arrows mark the inclusion bodies in mcoln1ab^−/− hair cells, and the asterisk marks inclusion bodies surrounding the hair cells. Scale bar, 1 μm. (C) and (D) Higher magnification showing more details of one of the hair cells in both WT (C) and mcoln1ab^−/− (D). Scale bar, 200 nm.

Figure 9

Autophagic dysfunction in mcoln1ab^−/− hair cells at 5 dpf. (A,B) Higher magnification of TEM showing autophagic vacuoles containing electro-dense inclusions, organelles and other cytoplasmic contents in mcoln1ab^−/− hair cells. (C) Typical lamellar structure in mcoln1ab^−/− hair cells. (D,E) Typical lamellar structures in MLIV fibroblasts. (F) Electron-transparent matrix and reduced cristae in mcoln1ab^−/− hair cell mitochondria. (G,H) Autolysosomes containing partially degraded mitochondria inside in mcoln1ab^−/− hair cells. Scale bar for (A) and (B), 200 nm. Scale bar for (C) to (H), 80 nm.

Figure 10

Lateral line neuromast structure of WT and mcoln1ab^−/− 5 dpf larvae by scanning electron microscopy (SEM). c, cupula, k, kinocilia; s, stereocilia. Scale bar, 10 μm.