UBAP1 mutations cause juvenile-onset hereditary spastic paraplegias (SPG80) and impair UBAP1 targeting to endosomes
概要
The aim of this study:
Hereditary spastic paraplegias (HSPs) are a clinically and genetically heterogeneous group of rare neurodegenerative disorders. To date, the genetic etiology remained unknown in approximately 40% of the HSP patients or families. Ubiquitin-associated protein 1 (UBAP1) is a subunit of mammalian endosomal sorting complex required for transport I (ESCRT-I). The importance of UBAP1 needs to be further underlined. In this study, we report six patients from four families with heterozygous truncating mutations in UBAP1 as the genetic cause of a new type of AD-HSP (SPG80). We created a mouse homologue of the mutation, and further characterized the UBAP1 wild-type (WT) and UBAP1 mutant by examining their functional properties in mouse hippocampus neurons.
Methods:
Six patients from four families were recruited and evaluated clinically by the Japan Spastic Paraplegia Research Consortium (JASPAC), the University of Yamanashi, and the University of Tokyo. All patients showed spastic paraplegia as a pure form of juvenile-onset HSP. We carried out whole-exome sequencing of genomic DNA from patients and both parents of patient III-7 in family #1. Genomic DNA was extracted from peripheral blood. We cloned mouse UBAP1 cDNAs and molecularly cloned mouse UBAP1-mutant cDNAs. Expression vectors for UBAP1 and ubiquitin (Ub) were constructed in the pCAII-EGFP and pCAII-HA vectors. Primary cultures of hippocampal neurons were prepared from eighteen-day-old pregnant Wistar rats. Immunostaining was performed at 13-15 DIV. Fluorescence images were acquired by confocal laser microscopy. We used the following primary antibodies: anti-MAP2 (1:400, GeneTex); anti-GFP (1:500, Life Technologies); anti-Rab5 (1:200, Thermo Fisher); anti-Rab7 (1:100, Thermo Fisher); and anti-HA (1:400, Roche Applied Science). To detect the immunofluorescence signals, Alexa Fluor 488- or 568- labeled secondary antibodies were used. For Western blotting, COS7 cells in 6-cm dishes were transfected with expression vectors using polyethylenimine.
The UBAP1 distribution pattern was morphologically classified by investigators blind as to the transfected construct into “punctate”, “homogeneous distribution”, and “others”. For each observation, these rates were calculated.
Results:
In family #1, we found only one likely candidate frameshift mutation (c.425_426del, p.K143Sfs*15) of the UBAP1 gene (GenBank: NM_016525.4). After analysis of family #1, we screened the whole-exome sequencing data for 2376 index cases including 330 cases with HSP from the JASPAC database and another 2046 cases to confirm that the UBAP1 gene is associated with HSP. We detected the same frameshift mutation (c.425_426del, p.K143Sfs*15) of the UBAP1 gene in IV-1 of family #2, another frameshift mutation (c.312delC, p.S105Pfs*46) in II-2 of family #3, and a nonsense mutation (c.535G>T, p.E179*) in III-2 of family #4 from the exome sequencing data. We reconfirmed all three variants of the UBAP1 gene on Sanger sequencing. Co-segregation was also confirmed. In protein function analysis, we demonstrated truncated cytoplasmic protein expression of UBAP1-mutant. This C-terminal deletion UBAP1-mutant of our disease model has lost its ubiquitin-binding domain, which is crucial for endosomal sorting of ubiquitinated cargos. The morphology on the overexpression of UBAP1-mutant is quite different from that of UBAP1-WT in a primary culture of mouse hippocampal neurons. The overexpression of UBAP1-WT generates discrete puncta in the neuron cytoplasm, which is absent from UBAP1-mutant neurons. The overexpression of UBAP1-mutant gave homogeneous diffusion in the neuron cytoplasm and throughout the nucleus. The full-length UBAP1protein is localized on expanded endosomes in the neuron cytoplasm. On the other hand, the UBAP1-mutant cannot be recruited to endosome membranes. The overexpression of UBAP1-WT is localized on expanded endosomes; we further demonstrated that UBAP1-WT interacts directly with ubiquitin on endosomes in vitro, while UBAP1-mutant loses its ability to bind ubiquitin as expected in silico.
Discussion:
We report that heterozygous loss of function mutations in UBAP1 lead to HSP (SPG80). The patients' common clinical features are juvenile-onset, slowly progressive, and a pure form of HSP with autosomal dominant or de novo inheritance. According to our statistics based on JASPAC data, HSP due to a UBAP1 mutation occurs at a frequency of 0.56% in all HSP patients and 1.5% in AD-HSP patients in Japan. In this study, we showed that the C-terminal deletion UBAP1-mutant of our disease model has lost its ubiquitin-binding domain and therefore has lost its ability to bind ubiquitin in vitro. The overexpression of UBAP1-WT interacts directly with ubiquitin on enlarged endosomes in mouse hippocampal neurons, while the UBAP1-mutant cannot be recruited to endosome membranes. Therefore, we conclude that UBAP1-mutant loses its function of endosomal sorting of ubiquitinated cargos in neurons. UBAP1 is coassembled with TSG101, VPS28, and VPS37A to form ESCRT-I via binding to the central stalk region of ESCRT-I. This mammalian ESCRT-I central stalk happens to be composed of the VPS37A-TSG101 combination. Hence, being the subunits of ESCRT-I, TSG101, VPS37A and UBAP1 may exhibit some similarities in the genotype to phenotype relationship. UBAP1 protein functions in the early endosomal compartments in neurons. Conversely, the UBAP1-mutant cannot be recruited to endosomes. Such a neuron-specific effect of loss of UBAP1 function perturbs endosomal fusion and sorting activities, thereby giving rise to the phenotype of a neurodegenerative disease such as HSP.
Conclusion:
We report for the first time mutations in UBAP1 gene as the genetic cause of a new type of pure AD-HSP. Functional studies reveal UBAP1 is involved in endosomal dynamics in vitro. Loss of UBAP1 function may perturb endosomal fusion and sorting of ubiquitinated cargos in neuron, which may serve as the pathogenesis in HSP. All the patients with novel heterozygous frameshift mutations in UBAP1 gene presented identical clinical features as a pure type of juvenile-onset AD-HSP. We propose to name this type of HSP Spastic Paraplegia-80 (SPG80). The knockout mice of our disease model are now being created. Further assays will be developed to enable the study in vivo, and models will be put in place to test new drug therapies.