ELOVL5 encodes for a multi-pass membrane protein involved in the elongation of long-chain polyunsaturated fatty acids.
At the time of writing, two missense variants have been associated to SCA38: c.214C>G – p.(Leu72Val), c.689G>T – p.(Gly230Val).
A third variant, c.779A>G – p.(Tyr260Cys) has been described in a Spanish pedigree. According to the gnomAD database (ver2.1), its frequency is 174/282674 (0.0006) in controls worldwide and 158/154164 (0.001) among European. These high frequencies strongly argue against a role of the c.779A>G variant in SCA38.
Of note, a nonsense c.304C>T – p.Gln102Ter variant has been reported in exon 4 of an alternative transcript of the ELOVL5 gene (NM_001242828): as this transcript has not been found expressed studying different human tissue, the variant should be considered intronic in the main transcript (NM_021814.4) and likely benign.
No pathogenic deletion/duplication has been reported.
The diagnosis of SCA38 is confirmed by sequencing the gene. If a family history of SCA38 is reported, and the mutation is known, direct testing for the pathogenic variant is recommended.
In other cases, ELOVL5 should be analyzed as part of a multigenic ataxia-related gene panel or of a full exome/genome sequencing. The interpretation of pathogenicity should be carefully evaluated especially considering the correct gene transcript, the frequency in the population, and the segregation in multiple members of the same family.
The molecular mechanism(s) causing SCA38 are uncertain: both loss-of-function and gain-of-function mechanisms have been proposed.