Characterization of novel SLC6A8 variants with the use of splice-site analysis tools and implementation of a newly developed LOVD database.

The X-linked creatine transporter defect is caused by mutations in the SLC6A8 gene. Until now, 66 synonymous and intronic variants in SLC6A8 were detected in our laboratory. To gain more insight in the effect of the detected variants, we applied five free web-based splice-site analysis tools to 25 published variants that were stratified as (non-)disease causing. All were correctly predicted to have no effect (n=18) or to cause erroneous splicing (n=7), with the exception of a pathogenic de novo 24 bp intronic deletion. Second, 41 unclassified variants, including 28 novel, were subjected to analysis by these tools. At least four splice-site analysis tools predicted that three of the variants would affect splicing as the mutations disrupted the canonical splice site. Urinary creatine/creatinine and brain MRS confirmed creatine transporter deficiency in five patients (four families), including one female. Another variant was predicted to moderately affect splicing by all five tools. However, transient transfection of a minigene containing the variant in a partial SLC6A8 segment showed no splicing errors, and thus was finally classified as non-disease causing. This study shows that splice tools are useful for the characterization of the majority of variants, but also illustrates that the actual effect can be misclassified in rare occasions. Therefore, further laboratory studies should be considered before final conclusions on the disease-causing nature are drawn. To provide an accessible database, the 109 currently known SLC6A8 variants, including 35 novel ones, are included in a newly developed LOVD DNA variation database.