Submission Details

KIF5A was first reported in relation to autosomal dominant inherited ALS in two separate studies both published in March 2018 (Brenner et al., PMID: 29342275 and Nicolas et al., PMID: 29566793). Per criteria outlined by the ClinGen Lumping and Splitting Working Group, we found difference in molecular mechanism and phenotypic variability of Hereditary Spastic Paraplegia and Charcot-Marie-Tooth disease type 2. Therefore, the following disease entities have been split into multiple disease entities, ALS 25 (OMIM:617921), Spastic paraplegia 10 (OMIM:604187) and Charcot-Marie-Tooth disease type 2. 42 variants (variant type, missense, frameshift and exon skipping) that have been reported in 42 probands in 8 publications (PMIDs: 29342275, 29566793, 30301576, 31422367, 31475037, 32815063, 32888732, 33077544) are included in this curation. More evidence is available in the literature, but the maximum score for genetic evidence (12 pts.) has been reached. Many of the variants reported in people with ALS are at the C-terminal end of the protein and involve either skipping of exon 27 or frameshift variants that cause the final 34 amino acids of the protein to be substituted for a novel sequence of 39 or more amino acids with a number of charged residues. This novel amino acid sequence appears to confer a toxic gain of function of the variant protein (PMIDs: 35430760, 35385738). Cellular functional evidence includes reduced survival of neurons, formation of toxic aggregates, mislocalisation of variant protein to the distal end of neurons, increased binding and activity on microtubules, lack of autoinhibition, and co-localisation and subsequent mislocalisation of wild-type protein. The evidence was considered strong and well replicated enough to upgrade the variant score to one for each first occurrence of a variant predicted to cause the tail domain toxic gain of function. This gene-disease relationship is also supported by experimental evidence, including expression assays (PMIDs: 24309898, 23006449) showing that KIF5A is expressed in healthy motor neurons and the levels are reduced in people with ALS. A model of C. elegans expressing human wild-type and exon 27 skipping KIF5A demonstrated neuronal loss and abnormal phenotype (PMID: 35430760). There are several mouse models of KIF5A knock-out in neurons (eg PMIDs: 12682084, 22466687), these were included in the curation but not scored as the mechanism is likely to be gain-of-function. In summary, KIF5A is definitively associated with autosomal dominant ALS. This has been repeatedly demonstrated in both the research and clinical diagnostic settings, and has been upheld over time.