Ubiquitous expression of dyrk1aa and dyrk1ab in the larval nervous system becomes confined to specific structures in the adult brain. RNA in situ hybridization in zebrafish larvae and adult brain sections with dyrk1aa (A, Aa, Ab, C, and E–G) and dyrk1ab (B, Ba, Bb, D, and H–J) specific probes and respective sense controls are shown in Fig. S1. dyrk1aa and dyrk1ab both are expressed throughout the central nervous system of 5 dpf zebrafish larvae (A and B). Transverse sections from different planes with strong expression for both paralogs in the entire gray matter (Aa, Ab, Ba, and Bb). Chromogenic mRNA in situ of adult brain sagittal sections for dyrk1aa (C) and dyrk1ab (D) with intense staining in the cerebellar granule cell layer and the torus longitudinalis. Fluorescent dyrk1aa (E–G and K–M) and dyrk1ab (H–J) in situ with subsequent ZebrinII immunohistochemistry confirms expression in the granule cell layer below the Purkinje cell layer but also reveals expression in Purkinje cells, which was further confirmed by confocal microscopy at cellular resolution in which dyrk1aa expression (K) colocalizes with ZebrinII expression in individual Purkinje cells (L and M, white arrows). dyrk1aa (E and K), and dyrk1ab (H) fluorescent mRNA in situ (green) followed by anti-ZebrinII antibody staining (F, I, and L; red) with overlay for dyrk1aa (G and M) and dyrk1ab (J). Black lines in panels A and B indicate the plane of sections shown in panels Aa, Ab, Ba, and Bb. Yellow and white arrowheads indicate granule cells (GCs) and Purkinje cells (PCs), respectively. The scale bars indicate 200 μm (A and B), 100 μm (Aa, Ab, Ba, Bb, and E–J), 500 μm (C and D), 100 μm (E–J), and 20 μm (K–M). Bc, bipolar cells; CCe, corpus cerebelli; Di, diencephalon; Dm, medial zone of the dorsal telencephalic area; GCL: granule cell layer; Ha, habenula; Hc, central zone of the dorsal telencephalic area; Hd, dorsal zone of the periventricular hypothalamus; Hy, hypothalamus; La, lobus caudalis; Me, metencephalon; ML, molecular layer; My, myelencephalon; PCL, Purkinje cell layer; TeO, tectum opticum; Th, thalamus; TL, torus longitudinalis; Va, valvula cerebelli; Vam, medial division of valvula cerebelli.

Purkinje cell–specific expression of human Dyrk1A in zebrafish.A, schematic presentation of the bidirectional construct for the generation of a stable transgenic line driving HA-tagged hDyrk1A expression in Purkinje cells. Two ca8 promoter-derived PC-specific enhancer element (cpce) dimers in opposite orientation (yellow triangles) are linked to E1B basal promoters on both sides directing the expression of mClover (right cistron) and HA-hDyrk1A (left cistron) to Purkinje cells. The construct is not drawn to scale. B–M, dorsal views (anterior to the left) onto the left and right cerebellar PC hemispheres (B) recorded by confocal microscopy. Transient transgenic 4 dpf larvae show expression of hDyrk1A, visualized by αHA antibody staining, and (C) FynmClover fluorescence in Purkinje cells. D, the merged image reveals coexpression of both proteins. Colocalization of the PC-specific αParvalbum antibody staining (E) and FynmClover (F) in the merged image (G) indicate that the hDyrk1A expression is confined to cerebellar PCs. H–M, whole-mount immunostaining on F2 larvae of a stable Tg(ca8-hDryk1A-mClover) line. H, maximal brightness projections of confocal microscopy image stack recordings after αHA antibody staining visualize hDyrk1A expression in the cerebellum. I, the corresponding mClover fluorescence labels PCs, whereas the merged image (J) demonstrates reliable coexpression of both proteins. Images at higher magnification display membrane-tagged mClover (K and K’) and nuclear/cytoplasmic hDyrk1A (L and L’) colocalized within the same cells (M and M’). White arrowheads indicate Purkinje cells. The scale bar represents 50 μm (B–M) and 20 μm (K’–M’). cpce, ca8 promoter–derived PC-specific enhancer element; dpf, days post fertilization; Dyrk1A, dual-specificity tyrosine phosphorylation–regulated kinase 1A; HB, hindbrain; PCs, Purkinje cells; TeO, tectum opticum.

hDyrk1A overexpression in PCs impairs morphology of the PC layer.A–H, PC layer of either PC-Dyrk1A larvae displaying green mClover fluorescence or PC-RFP control larvae with red fluorescence between 4 and 7 dpf. Dorsal view (anterior to the left) of left and right PC hemispheres of different PC-Dyrk1A larvae from the same developmental stage with either strong (A–D) or mild (A’–D’) phenotype in PC layer morphology and corresponding controls (E–H). I–M, quantification of morphological abnormalities from PC-RFP control (red) and PC-Dyrk1A (blue) larvae between 4 and 7 dpf (n = 12–19). Measurements of anterior (I) and posterior (J) distance between the two PC layer hemispheres, anterior–posterior length (K) and area (L) of individual hemispheres, as well as the numbers of PCs (M) per hemisphere. I–M, data are the mean ± SD (error bars). Statistical analysis was performed using two-way ANOVA, followed by Tukey's post hoc multiple comparisons test. I, interaction: F = 10.25, p < 0.0001; row factor: F = 1.725, p = 0.1642; column factor: F = 59.45. p < 0.0001. ∗p = 0.0437, ∗∗p = 0.0051, ∗∗∗p = 0.0004, and ∗∗∗∗p < 0.0001. J, interaction: F = 1.682, p = 0.1732; row factor: F = 0.8437, p = 0.4719; column factor: F = 59.45, p < 0.0001 and ∗∗∗∗ p < 0.0001. K, interaction: F = 4.523, p = 0.0045; row factor: F = 56.01, p < 0.0001; column factor: F = 80.41, p < 0.000. ∗p = 0.0272, ∗∗∗ RFP control 4 dpf versus 5 dpf RFP p = 0.002, ∗∗∗PC-Dyrk1A 5 dpf versus 6 dpf p = 0.0003, ∗∗∗∗ p < 0.0001. L, interaction: F = 4.035, p = 0.0081; row factor: F = 31.15, p < 0.0001; column factor: F = 56.54, p < 0.0001. ∗ PC-Dyrk1A 5 dpf versus 6 dpf p = 0.0486; ∗ RFP control versus PC-Dyrk1A 6 dpf p = 0.0272, ∗ RFP control 6 dpf versus 7 dpf p = 0.0104, ∗∗∗∗ p < 0.0001. M, interaction: F = 0.4441, p = 0.7221; row factor: F = 32.67, p < 0.0001; column factor: F = 6.727, p = 0.0113. ∗ RFP control 5 dpf versus 6 dpf 0.0321, ∗∗ PC-Dyrk1A 5 dpf versus 6 dpf p = 0.0017. N, schematic drawing to illustrate the positions of measurements quantified in panels I–K. Anterior to the left (A–H) and to the top (N), respectively. The scale bar represents 50 μm (A–H and N). ap, anterior–posterior length; da, distance anterior; dp, distance posterior; dpf, days post fertilization; Dyrk1A, dual-specificity tyrosine phosphorylation–regulated kinase 1A; PCs, Purkinje cells.

hDyrk1A overexpression in PCs compromises swimming behavior.A–D, swimming activity of 5 dpf WT and heterozygous PC-Dyrk1A larvae; mean values ±SD of 30 min tracks. A, PC-Dyrk1A larvae swam significantly shorter distances than WT larvae (two-tailed unpaired t test, p = 0.022) and B, consequently displayed a reduced average swim speed (two-tailed unpaired t test, p = 0.0041). C, the distance swum with slow movements (0.2< slow speed <21 mm/s) was significantly decreased (two-tailed unpaired t test, p = 0.0228) in PC-Dyrk1A larva (D) as was the distance covered by fast movements (fast speed >21 mm/s, nonparametric t test p = 0.0385). The number of movements was significantly decreased in PC-Dyrk1A larva (E, two-tailed unpaired t test, p = 0.0012), whereas the period of activity was almost equal in both groups (F, nonparametric t test, ns, p = 0.0612). All values are the means ± SD (error bars). WT: n = 44; PC-Dyrk1A: n = 50; ∗p < 0.05; ∗∗p < 0.02; ∗∗∗p < 0.005. dpf, days post fertilization; Dyrk1A, dual-specificity tyrosine phosphorylation–regulated kinase 1A; ns, not significant; PCs, Purkinje cells.

A compromised PC layer in adult PC-Dyrk1A brains is accompanied by a strong loss of synaptic puncta. A–C, the expression of Dyrk1A and mClover persists in adult brains of PC-Dyrk1A transgenic fish as shown on frozen cerebellar sagittal sections stained with (A) αHA or (B) mClover expression. C, the overlay of the images documents the coexpression of hDyrk1A and the fluorescent mClover protein. dyrk1aa in situ hybridization (D) with DAPI counterstaining (E) demonstrated that the endogenous expression of zebrafish dyrk1aa is maintained in the cerebellum of adult PC-Dyrk1A fish. F–K, PC-specific ZebrinII immunohistochemistry with DAPI counterstaining on adult frozen sagittal cerebellar sections of PC-Dyrk1A (F–H) and PC-RFP controls (I–K). Purkinje cells are no longer found in an orderly fashion between granule cell layer (GCL) and the molecular layer (ML) in PC-Dyrk1A fish as in controls but appear scattered throughout the ML (A–C, white arrowheads). This is confirmed by quantitative analysis of (L) cell nucleus density in the ML and (M) the distance of PCs to the GCL (unpaired two-tailed t test, p < 0.0001). Compared with controls (N and P), the synaptic density in the molecular layer was reduced in PC-Dyrk1A brains (O and Q) shown by anti-Synaptophysin immunohistochemistry counterstained with DAPI and quantified by counting the density of synaptic puncta (R) (unpaired two-tailed t test, p < 0.0001). Values are the means ± SD (error bars). Anterior is to the left and dorsal up. The scale bar represents 100 μm (A–K), 20 μm (insets F and I), 20 μm (N and O), and 5 μm (P and Q). DAPI, 4’,6-diamidino-2-phenylindole; GCL, granule cell layer; ML, molecular layer. ∗∗∗∗p < 0.0001.

Novel tank diving test reveals anxiety-like behavior in adult Tg(Car8-hDryk1A-mClover) fish.A–F, swimming behavior of adult WT (n = 14) and PC-Dyrk1A fish (n = 10) examined in the novel tank diving test. The number of zone transitions (A, two-tailed unpaired t test, p = 0.003), the time in the top part (B, nonparametric t test, p = 0.0015), the distance covered (C, two-tailed unpaired t test, p < 0.0001), and the speed (D, two-tailed unpaired t test, p < 0.0001) were determined in 6-min recordings. All parameters analyzed were significantly reduced in PC-Dyrk1A fish. E and F, representative tracks of a control (E) and PC-Dyrk1A zebrafish with bottom (light blue) and top part (light gray) of the tank. Means ± SD. The scale bar represents 3 cm. Dyrk1A, dual-specificity tyrosine phosphorylation–regulated kinase 1A; PCs, Purkinje cells. ∗∗p < 0.01, ∗∗∗∗p < 0.0001.

Dryk1A inhibitors rescue the morphological Purkinje cell layer impairments of PC-Dyrk1A larvae.A, scheme of inhibitor administration. Inhibitors dissolved in DMSO were diluted to the respective concentrations in 30% Danieau. 3 dpf larvae were treated for 4 days, and the inhibitor was changed daily. In vivo imaging of PC hemispheres of all groups was performed between 2 and 4 dpt. B–D, measurements of the anterior distance (μm) of the PC hemispheres for (B) ProINDY, (C) Leucettine L41, and (D) KuFal194 treated larvae. Each inhibitor test encompasses the following groups: PC-RFP control larvae treated with DMSO (Ctrl, DMSO), PC-RFP control larvae treated with inhibitor (Ctrl + inhibitor), PC-Dyrk1A larvae treated with DMSO (Dyrk1A DMSO), and PC-Dyrk1A larvae treated with inhibitor (Dyrk1A + inhibitor). B–D, data are the mean ± SD (error bars). Statistical analysis was performed using three-way ANOVA, followed by Tukey's post hoc multiple comparisons test. B, row factor: F = 0.9372 p = 0.3953, row factor (Ctrl DMSO Ctrl PI versus Dyrk1A DMSO Dyrk1A PI): F = 0.9372, p = 0.3953; row factor (Ctrl DMSO Dyrk1A DMSO versus Ctrl PI Dyrk1A PI): F = 10.47, p = 0.0017; row factor x (Ctrl DMSO Ctrl PI versus Dyrk1A DMSO Dyrk1A PI): F = 12.75, p < 0.0001; row factor x (Ctrl DMSO Dyrk1A DMSO versus Ctrl PI Dyrk1A PI): F = 4.843, p = 0.0099; row factor (Ctrl DMSO Ctrl PI versus Dyrk1A DMSO Dyrk1A PI) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl PI Dyrk1A PI): F = 14.47, p = 0.0002; row factor x (Ctrl DMSO Ctrl PI versus Dyrk1A DMSO Dyrk1A PI) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl PI Dyrk1A PI): F = 3.555, p = 0.0324. ∗∗∗p = 0.0003 and ∗∗∗∗p < 0.0001. C, row factor: F = 0.1570, p = 0.8549; row factor (Ctrl DMSO Ctrl L41 versus Dyrk1A DMSO Dyrk1A L41): F = 30.85, p < 0.0001; row factor (Ctrl DMSO Dyrk1A DMSO versus Ctrl L41 Dyrk1A L41): F = 2.139, p = 0.1466; row factor x (Ctrl DMSO Ctrl L41 versus Dyrk1A DMSO Dyrk1A L41): F = 9.839., p = 0.0001; row factor x (Ctrl DMSO Dyrk1A DMSO versus Ctrl L41 Dyrk1A L41): F = 2.441, p = 0.0921; row factor (Ctrl DMSO Ctrl L41 versus Dyrk1A DMSO Dyrk1A L41) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl L41 Dyrk1A L41): F = 3.689, p = 0.0575; row factor x (Ctrl DMSO Ctrl L41 versus Dyrk1A DMSO Dyrk1A L41) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl L41 Dyrk1A L41): F = 1.708, p = 0.1863. ∗∗Dyrk1A DMSO 2 dpt versus 4 dpt p = 0.0068; ∗∗Dyrk1A DMSO versus Dyrk1A L41 4 dpt p = 0.0065; ∗∗∗∗ p < 0.0001. D, row factor: F = 1.843, p = 0.1617; row factor (Ctrl DMSO Ctrl KuFal194 versus Dyrk1A DMSO Dyrk1A KuFal194): F = 108.3, p < 0.0001; row factor (Ctrl DMSO Dyrk1A DMSO versus Ctrl KuFal194 Dyrk1A KuFal194): F = 10.35, p = 0.0016; row factor x (Ctrl DMSO Ctrl KuFal194 versus Dyrk1A DMSO Dyrk1A KuFal194) F = 3.088, p = 0.0482; row factor x (Ctrl DMSO Dyrk1A DMSO versus Ctrl KuFal194 Dyrk1A KuFal194): F = 0.5474, p = 0.5795; row factor (Ctrl DMSO Ctrl KuFal194 versus Dyrk1A DMSO Dyrk1A KuFal194) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl KuFal194 Dyrk1A KuFal194): F = 23.42, p < 0.0001; row factor x (Ctrl DMSO Ctrl KuFal194 versus Dyrk1A DMSO Dyrk1A KuFal194) x (Ctrl DMSO Dyrk1A DMSO versus Ctrl KuFal194 Dyrk1A KuFal194): F = 2.762, p = 0.0661. ∗Ctrl DMSO versus Dyrk1A DMSO 2 dpt p = 0.0120; ∗Dyrk1A DMSO versus Dyrk1A KuFal 10 μM 3 dpt p = 0.0111; ∗∗p = 0.0013; ∗∗∗∗p < 0.0001. E–N, representative images of PC hemispheres of reconstructed confocal laser scanning z-stacks for (E and J) PC-RFP DMSO controls, (F and K) PC-Dyrk1A DMSO-treated larvae, (G and L) ProINDY-treated larvae, (H and M) Leucettine L41-treated larvae, and (I and N) KuFal194-treated larvae. E–I, 2 dpt and (J–N) 4 dpt. The scale bar represents 50 μm (E–N). dpt, days post treatment; DMSO, dimethyl sulfoxide; Dyrk1A, dual-specificity tyrosine phosphorylation–regulated kinase 1A; PCs, Purkinje cells.