Mutation in the Scyl1 gene encoding amino-terminal kinase-like protein causes a recessive form of spinocerebellar neurodegeneration

Abstract

Here, we show that the murine neurodegenerative disease mdf (autosomal recessive mouse mutant 'muscle deficient') is caused by a loss-of-function mutation in Scyl1, disrupting the expression of N-terminal kinase-like protein, an evolutionarily conserved putative component of the nucleocytoplasmic transport machinery. Scyl1 is prominently expressed in neurons, and enriched at central nervous system synapses and neuromuscular junctions. We show that the pathology of mdf comprises cerebellar atrophy, Purkinje cell loss and optic nerve atrophy, and therefore defines a new animal model for neurodegenerative diseases with cerebellar involvement in humans.

Figure 1

A single thymidine-nucleotide insertion causes disruption of Scyl1 expression in the mdf−mouse. (A) Mapping of sequence tagged sites located mdf to an approximately 0.9 Mb chromosomal region between D19Mit59 and D19Mit109. Location of Scyl1 is indicated and the exon–intron structure depicted (the exons are numbered). The mdf mutation is indicated by an arrow (inset: representative sequence traces showing the nucleotide insertion at position c.1169_1170insT). (B) Schematic ‘in silico' annotation of Scyl1 protein domains (numbers indicate amino-acid (aa) residues): serine–threonine-tyrosine protein kinase, HEAT repeats, secretory carrier membrane protein domain (KOG3088). Positions of peptides for raising antibodies (Ab) are shown. The premature stop codon created by the mdf mutation is indicated (arrowhead). (C) Quantitative reverse transcription–PCR of total RNA isolated from skeletal muscle (Skm) or brain (Br) showing a marked reduction of Scyl1 messenger RNA in Scyl1^mdf/mdf. Error bars indicate standard deviation (n=4). (D) Western blot loaded with extracts from brain (Br), cerebellum (Cb), spinal cord (Sc) and skeletal muscle (Skm) of wild-type Scyl1^+/+ and Scyl1^mdf/mdf mice probed with polyclonal antibody raised against the amino terminus of Scyl1. In wild-type tissues, the antibody detected an approximately 105 kDa band (arrowhead) that was completely absent in corresponding tissues from Scyl1^mdf/mdf mice. A second, minor reactive band of approximately 70 kDa is also indicated. MW, molecular weight standard (∼70, ∼100, ∼130, ∼150 kDa). (E) Detection of Scyl1-mRNA by in situ hybridization (ISH) in a wild-type mouse brain section showing high expression in the granular and Purkinje cell layers of the cerebellum (gcl), the hippocampus (hc) and the gyrus dentatus (gd). Right panel is a magnified view showing abundant Scyl1 expression in the synaptic region of the cerebellar granular layer. Left panel: control experiments showing specificity of the antisense ISH probe as opposed to the sense control probe.

Figure 2

Scyl1 protein is a neuronal protein expressed in the axoplasm and concentrated at synapses. (A,B) Immunodetection of Scyl1 on semicoronal brain sections from (A) wild type Scyl1^+/+, but a lack of reactivity in (B) Scyl1^mdf/mdf. Note the strong Scyl1-specific immunosignal of the neuropil (A, white stars) and of the choroid plexus (A, arrow) in wild type. The neuropil in mutant brain does not yield comparable immunosignals (B, white stars). ccl, corpus callosum. (C) Prominent Scyl1 staining of cortical neurons (arrows) and (D) brain-stem neurons (arrows) of wild-type mice. (E) Prominent Scyl1 reactivity of anterior horn motor neurons (arrows) in wild-type spinal cord. (F) Confocal immunofluorescence images showing intense Scyl1 staining in the cytoplasm (arrowheads) and axons (arrows) of murine cerebral cortex neurons (white star, nonreactive nucleus). (G) Section from aged human brain showing SCYL1-specific immunosignal accumulation within axonal spheroids (arrows). (H) Strong axoplasmic Scyl1-specific staining of an intramuscular nerve branch in wild-type skeletal muscle (skm; arrowheads in inset). Scyl1 reactivity of a neuromuscular junction (NMJ) is indicated (arrow). (I) Scyl1 distribution at a wild-type NMJ: pronounced labelling of presynaptic terminal axons (arrows) and weaker reactivity of the post-synaptic sarcoplasm (arrowhead). α-Bungarotoxin (BgTx, arrowhead) was used as a postsynaptic marker. Confocal overlay of Scyl1 immunosignals (green) with BgTx-specific red fluorescence shows predominant presynaptic Scyl1 expression, which was not detectable at NMJs in Scyl1^mdf/mdf mice (lower panel). All immunostainings shown were performed using the antibody raised against the Scyl1 C terminus.

Figure 3

Purkinje cell degeneration and cerebellar atrophy in mdf. (A) Immunodetection of SCYL1 in Purkinje cell (PC) somata (arrows) in human cerebellum with fine granular immunosignal accumulation in the synaptic region of the granular layer (white stars). (B) Scyl1-immunolocalization (carboxy-terminal antibody) in wild-type (Scyl1^+/+) mouse (255 days) cerebellum showing strong labelling within the PC layer. Arrows indicate the periodic continuity of PCs in wild-type cerebellum. (C) By contrast, PCs (arrows) in Scyl1^mdf/mdf-cerebellum (age-matched) were frequently abnormally shaped and displayed basophilic, condensed cytoplasm (H&E-staining). Arrowheads indicate discontinuity of the PC layer owing to the loss of single cells. Inset: two remaining PCs neighboured by persisting basket formations devoid of PCs (arrowheads), as shown by Bielschowsky staining. (D) Quantification of PC loss in mdf. PC counts in various areas of wild-type and Scyl1^mdf/mdf cerebella show an overall loss of 42% in young mice (P=5.638e−05, Kruskal-Wallis rank sum test) and 48% in adult mice (P=1.539e−07), indicating early-onset and slowly progressive PC loss. Notches indicate confidence intervals of medians. (E) Immunostaining of calbindin D-28k in the cerebella of 274-day-old wild-type and (F) Scyl1^mdf/mdf mice, showing pronounced loss of PCs (arrowheads). Insets: commensurate close-up views of representative PCs showing regular and complex dendritic arbors in wild-type-cerebella (inset in E), whereas multiple swellings (inset in F, arrowheads) and paucity of arborization are hallmarks in mdf. (G,H) H&E-stained median-sagittal sections of cerebella from (G) age-matched wild-type and (H) Scyl1^mdf/mdf mice showing atrophy of mdf cerebellum compared with wild-type. Scale bars, 1 mm. H&E, haematoxylin and eosin.