Ubap1 knock-in mice reproduced the phenotype of SPG80

Generation of Ubap1 knock-in mice

UBAP1 comprises a 502 amino acid residue protein that is widely expressed in human tissues and broadly conserved among species, and human UBAP1 shows more than 90% sequence identity to the mouse Ubap1 protein. Mammalian UBAP1 consists of three major domains: the N-terminal UMA (UBAP1-MVB12-associated) domain, the HD-PTP (HIS-domain protein tyrosine phosphatase) binding region, and the C-terminal SOUBA (solenoid of overlapping ubiquitin-associated) domain [8] (Fig. 1B). The mutations of UBAP1 in the four families we previously reported as well as the other families with SPG80 reported from all over the world so far all result in prematurely truncated proteins with complete loss of the C-terminal SOUBA domain, the N-terminal UMA domain remaining intact [4,5,6, 18,19,20]. Therefore, we generated knock-in mice in which the SOUBA domain was completely deleted with the UMA domain being unaffected. In accordance with the four families reported previously [6], we generated mutant mice genetically similar as to the mutation (c.535 G > T, p.E179*) that resulted in the SPG80 phenotype. The mutation in the knock-in mice obtained was c.527dupA (p.E176Efx23), which would theoretically cause the same dysfunction (Fig. 1A, B, C). However, it is noteworthy that CRISPR can induce off-target editing at genomic positions that imperfectly match the sgRNA sequence. Therefore, given the possibility that there are genomic regions of close homology to the sgRNAs used for Cas9-dependent cleavage, the possibility that these off-target sites are also modified should be considered. We screened off-target candidates from the whole mouse genome sequence using CRISPRdirect (https://crispr.dbcls.jp/) (Supplementary Table 1). Examination of the mouse genome DNA with 20-nucleotide gRNA sequences revealed no 15 or more nucleotide matches, but three sequences with 14-nucleotide matches. These three off-target candidates were examined by direct sequencing analysis. No mutations were found in any of the off-target candidates, indicating that the chance that Cas9-mediated off-target mutation had occurred was low.

Fig. 1figure 1

A, B UBAP1 gene and protein structure, and predicted effects on the protein structure of the mutation in our UBAP1+/E176Efx23 knock-in mouse model. The locations of the mutations in our patient (c.535 G > T, p.E179*) and the model mouse (c.527dupA, p.E176Efx23) we created are shown. The locations of the PCR primers used to screen for the presence of the targeted mutation are salso hown (black arrow). C Sanger sequencing results for wild-type mice and Ubap1+/E176Efx23 knock-in mice. The mutation of c.527dupA (red arrow) was detected in Ubap1+/E176Efx23 knock-in mice, while no mutation was detected in wild-type mice

Ubap1+/E176Efx23 knock-in mice were delivered in expected Mendelian ratios and were viable with normal reproduction. However, none of the Ubap1E176Efx23/E176Efx23 mice were born, which is consistent with previous finding that Ubap1 is expressed in a wide range of tissues, and its loss in homozygous mice is embryonic lethal [7].

mRNA expression of UBAP1 mutant allele in brain

To determine whether the observed truncating variants would lead to nonsense-mediated mRNA decay in mouse brains, we evaluated the mRNA expression of the mutant allele. RT-PCR was performed on RNA extracted from the brain tissue from the wild-type and heterozygous Ubap1+/E176Efx23 mutant knock-in mice, and the RNA was sequenced by the Sanger method. Surprisingly, the heterozygous c.527dupA p.E176Efx23 mutant transcript was detected in the mutant mice cDNA, indicating escape from nonsense-mediated mRNA decay (Fig. 2A). To confirm the ratio between wild-type and mutant mRNA, we performed RFLP analysis on mice brain cDNA. The amplified cDNA from the heterozygous Ubap1+/E176Efx23 mutant was digested with MboII (Fig. 2B). We observed two fragments of 91 bp and 68 bp (23 bp fragment not visible), indicating the mutant mRNA was expressed in the brain (Fig. 2C). The intensity of the observed fragments is approximately half that of the fragments observed in the knock-in mice without the restriction enzyme, indicating that the mutant mRNA is normally expressed.

Fig. 2figure 2

A RT-PCR of brain tissue from wild-type and heterozygous Ubap1+/E176Efx23 mutant knock-in mice at 7 months of age indicates that truncated UBAP1 mRNA escapes nonsense-mediated decay in mutant mice brains. The boted area indicates the location of c.527dupA. B Schematic representation of RFLP analysis. Primers used in RT-PCR and their positions are indicated by arrows (upper). This site shows the expected restriction fragments following digestion with MboII. C Electrophoresis pattern of RFLP. Lanes 1, 2, and 3 shows the 91 bp fragment, and lane 4 shows both the 91 bp and 68 bp fragments (fragment 23 not visualized)

We next performed immunoblot analysis to evaluate both the wild-type and potential truncated mutant Ubap1. Although we thoroughly validated several antibodies raised against the N-terminal region of mouse Ubap1, unfortunately, we failed to detect endogenous Ubap1 in the mouse brain. For this reason, we prepared a plasmid expressing the N-terminally Flag epitope-tagged Ubap1 mutant, and then validated its expression in Neuro2a cells by transfection. After transfection, we successfully detected both Ubap1 wild-type and Ubap1 mutant expression in Neuro2a cells on Western blotting (Supplementary Fig. 1A). This suggests the mutant protein could be normally expressed.

Ubap1 +/E176Efx23 knock-in mice developed progressive spastic gait

Ubap1+/E176Efx23 knock-in mice appeared normal at birth and showed no obvious difference in body weight compared to the wild-type (Fig. 3A). At the age of two months, there were no significant differences in the numbers of slips and falls between Ubap1+/E176Efx23 knock-in mice and Ubap1 wild-type mice. However, at the age of three months, the numbers of slips and falls were significantly increased in Ubap1+/E176Efx23 knock-in mice. On the other hand, the wild-type mice showed fewer slips and falls during the test (Fig. 3B, C). These increases in frequency were thought to be correlated with decreases in motor coordination of the mutant mice. They also showed a marked decrease in the ability to grasp sticks with the hind limbs, which is similar to the phenotype reported for other HSP mouse models (Fig. 3B) [21]. At 6 months of age, they had difficulty crossing a beam and showed significant differences in gait from wild-type mice (Supplementary Videos 1 and 2).

Fig. 3figure 3

Ubap1 knock-in mice developed progressive spastic gait. A There was no difference in body weight between the genotypes at 3, 5, and 7 months of age (WT: N = 16, KI: N = 13; Welch’s t-test; n.s. not significant). B Representative photos of wild-type and UBAP1+/E176Efx23 knock-in mice walking on a beam, taken from the side. Ubap1+/E176Efx23 knock-in mice show a marked decrease in the ability to grasp the beam with their hind limbs (compare the dashed circles for the two genotypes). C Ubap1+/E176Efx23 knock-in mice showed an apparent age-dependent increase in the number of slips on the beam (WT: N = 16, KI: N = 13; Mann–Whitney U test; n.s. not significant, **: p < 0.01). D The foot-base-angle (FBA) at the toe-off position of the hind leg is indicated by the white line (7 months old). E There is a significant age-dependent decrease in the FBA of Ubap1+/E176Efx23 knock-in mice (WT: N = 13, KI: N = 15; Welch’s t-test; **p < 0.01, ***p < 0.001). Results are expressed as means ± SD

At 3 months of age, there was also a significant decrease in the FBA compared to in the wild-type (Fig. 3D, E). At three months, the FBA had decreased to about 65 degrees in mutant mice, and then it decreased progressively with growth to about 50 degrees at six months. The same type of gait disturbance was observed in female mice as well as male ones. No gender differences in human SPG80 patients were found in the previous studies [4,5,6, 18,19,20], which is consistent with the results of our animal study. No physical change other than the gait disturbance was observed, and the combination of hindlimb muscle weakness and spasticity closely resembled the clinical symptoms of human HSP patients, suggesting that Ubap1+/E176Efx23 knock-in mice can be a clinically valid disease model for SPG80.

Mutations in Ubap1 result in axonopathy of the central nervous system

Axonal degeneration has been reported as the main pathological change in other pure-type HSP mouse models, such as SPG4 and SPG31 [22, 23]. Therefore, we performed toluidine blue staining of transverse semi-thin sections of the spinal cord of Ubap1+/E176Efx23 mice at seven months of age to observe axons. The characteristic change in Ubap1+/E176Efx23 knock-in mice was the selective loss of thick myelinated fibers close to the spinal cord surface (Fig. 4A). In wild-type mice, myelinated fibers distributed bimodally with diameters of approximately 7 μm and 10 μm were most abundant, whereas in Ubap1+/E176Efx23 knock-in mice, myelinated fibers were destributed unimodally, with as small as 7 μm fibers being most abundant (Fig. 4B, C), which is indicative of axonal degeneration. We also observed direct evidence of axonal degeneration, which has also been observed in other pure-type HSP mouse models (Fig. 4D).

Fig. 4figure 4

Heterozygous deletion of Ubap1 results in axonopathy of the central nervous system. A Semithin sections of the corticospinal tract in the spinal cords of 7-months-old mice at the lumbar level. Scale bar, 20 μm. B The sizes of axons were determined and the percentage of each size is shown. C The violin plots show that thick myelinated fibers are selectively lost in Ubap1+/E176Efx23 knock-in mice (n = 216 each; Mann–Whitney U test; ***p < 0.001). D Arrows denote direct evidence of axonal degeneration, which was not observed in the wild-type mice. Scale bar, 20 μm

Mutations in Ubap1 affect the distributions of Rab5 and Rab7

ESCRT-I plays a role in the recognition of ubiquitinated cargoes in endosomes and the transport of these cargoes to late endosomes to form MVBs [24]. Because the UBAP1 protein functions in the early endosomal compartment of neurons [18], brain and spinal cord sections from 7-month-old mice of both genotypes were collected by the same method as that described above, immunostained with Rab5 and Rab7, and then photographed under confocal microscopy. Rab5 is localized to the early endosomes, which regulates the vesicular trafficking and the fusion of early endosomes during endocytosis [25]. Rab7 is localized to late endosomes and lysosomes, where it regulates vesicular trafficking and membrane fusion processes [26, 27].

Most Rab5 and Rab7-positive endosomes are constant in size, and are distributed near the plasma membrane and around the nucleus, respectively, [28, 29]. In our study, both the two markers (Rab5 and Rab7) showed similar fluorescence patterns for each genotype, and the results were reproducible. In wild-type mice, they were localized around the nucleus, and in Ubap1+/E176Efx23 knock-in mice, they were diffuse in the neurons (Fig. 5A). To quantify these findings, we performed nearest neighbor analysis and confirmed that there is a difference in distribution between the wild-type and Ubap1+/E176Efx23 knock-in mice (Fig. 5B).

Fig. 5figure 5

Mutations in Ubap1 affect the distributions of Rab5 and Rab7. A Confocal images of spinal cord sections from 7-month-old mice stained for Rab5 (green), Rab7 (green), and DAPI (blue). Scale bar, 10 μm. B Nearest neighbor analysis was performed to quantify the distributions of Rab5 and Rab7 puncta, which are shown in the graph (Rab5: n = 13, Rab7: n = 12; Mann–Whitney U test; ***: p < 0.001). C, D Quantification of the number and size of Rab5 and Rab7 positive puncta was performed (Rab5: n = 13, Rab7: n = 12; Mann–Whitney U test; ***: p < 0.001)

In addition to this distribution pattern change, Ubap1+/E176Efx23 knock-in mice showed smaller endosomes and an increase in their vesicle number (Fig. 5C, D). Alterations in the structure and distribution of Rab proteins are closely related to disease onset and pathogenesis, suggesting that the abnormal endosome formation and fusion observed in Ubap1+/E176Efx23 knock-in mice would lead to impaired vesicle traffic, which contributes to the disease pathogenesis [30,31,32].

Mutation of Ubap1 results in accumulation of ubiquitinated proteins and neuron loss in the spinal cord

To assess the effects of the Ubap1 mutant on the central nervous system, 7-month-old wild-type mice (N = 4) and Ubap1+/E176Efx23 knock-in mice (N = 3) were perfused with paraformaldehyde, and then their brains and spinal cords were extracted. UBAP1 is one of the subunits of ESCRT-I and plays a role in the sorting of ubiquitinated cargoes by binding to ubiquitin via its SOUBA domain [8]. Since the UBAP1-mutant has lost its ability to bind ubiquitin, it is expected that the sorting of ubiquitinated cargoes is severely affected. Therefore, we examined ubiquitin and autophagosome LC3 in neurons of the lumbar spinal cord by immunostaining. Compared within wild-type mice, the drastic accumulation of ubiquitinated proteins was observed in Ubap1+/E176Efx23 knock-in mice (Fig. 6A, B). In addition, autophagosome accumulation co-localized with the aggregation of ubiquitinated proteins was detected (Fig. 6A), indicating dysfunction of autophagy and the endosomal-lysosomal pathway. Accumulation of ubiquitinated proteins was also observed on Western blotting of proteins extracted from brains of Ubap1+/E176Efx23 knock-in mice (Fig. 6C, D), and Neuro2a cells with exogenous expression of Flag-UBAP1E176Efx23 (Supplementary Fig. 1B, C).

Fig. 6figure 6

Mutation of Ubap1 results in the accumulation of ubiquitinated proteins and loss of neurons in the spinal cord. A Confocal images of spinal cord sections from 7-month-old mice stained with ubiquitin (red), LC3 (green), and DAPI (blue). Yellow in the merged images indicates colocalization of ubiquitin and LC3. Scale bar, 10 μm. B Comparison of the area stained with ubiquitin in a single neuron is shown in the graph (WT: n = 28, KI: n = 30; Mann–Whitney U test; ***: p < 0.001). C, D Quantification of ubiquitinated proteins from mouse spinal cord tissue by Western blotting. The expression of GAPDH is shown as a loading control. Mutations in Ubap1 result in increases in the ubiquitinated proteins (WT: N = 4, KI: N = 4; Welch’s t-test; *: p < 0.05). E Confocal images of spinal cord sections from 7-month-old mice stained for NeuN (green). The image on the right is a higher magnification of the area indicated by the square. Scale bar, 100 μm. F Quantification was performed. Mutations in UBAP1 reduced the number of neurons in the spinal cord (WT: n = 14, KI: n = 11; Mann–Whitney U test; *: p < 0.05). G Mutations in Ubap1 caused an increase in the size of the anterior horn cells of the spinal cord (WT: n = 55, KI: n = 42; Mann–Whitney U test; ***: p < 0.001)

To assess the effects of the Ubap1 mutant on neuron degeneration, sections of the spinal cords were labeled by NeuN-immunostaining. The results showed that the number of neurons in the spinal cord of Ubap1+/E176Efx23 knock-in mice was reduced compared to in the wild-type mice at seven months of age (Fig. 6E, F). Notably, there was apparent hypertrophy of the anterior horn cells (Fig. 6E, G). Dysfunction of autophagy or the endosomal-lysosomal pathway might result in neuronal hypertrophy and a decrease in the number of neurons.

留言 (0)

沒有登入
gif