Figure 1

COQ2 variants and transcripts. The upper three panels in blue represents the protein coding transcripts of the COQ2 gene and the exons are numbered (transcripts which cause nonsense mediated decay or transcripts which produce no protein are not shown). Transcripts and protein domains are referred to Ensemble (http://useast.ensembl.org/index.html) and 1000 Genomes (http://www.1000genomes.org/home). The lower three panels represent the protein domains and motifs by Forsgren and colleagues [14], TIGRFAMs database (http://www.tigr.org/TIGRFAMs) [15] and Pfam database (http://pfam.xfam.org/) [16]. There were no transcripts which start from second ATG. Red box indicates the allylic polyprenyl diphosphate substrate-binding site. Purple Boxes indicate the six transmembrane domains (TMD) predicted. The orange box and green box show the areas of 4-hydroxybenzoate polyprenyl transferase (Accession: TIGR01474) and UbiA prenyltransferase family (Accession: PF01040) respectively. Potential risk variants and p.R22X shown in black were identified in multiple system atrophy (MSA) [4, 13, 17]. p.R22X only exists on the longest isoform, COQ2-001, and does not associated with MSA. ^†ENST00000439031 was missing in the latest Ensembl release 77 - October 2014. *Asterisks indicate variants found in our MSA series. Arrow shows an intronic variant, c403 + 10G > T (Exon1 + 10). Mutations in blue were identified in primary coenzyme Q10 (CoQ10) deficiency [5–12]. Mutations in red (p.S146N, p.R197H and p.R387X) were identified both in MSA and primary CoQ10 deficiency [4, 6, 8, 13].