We have performed thorough screening of
DJ-1 to detect mutations which may cause PD. Rather than screening our entire sample of multiplex PD families, we have employed a more efficient approach that utilized data from our previous genome screen to prioritize families for testing based upon their family specific LOD score at the
DJ-1 locus. We did not identify any gene dosage alterations in our 93 PD cases. We identified only one missense mutation, which has been shown in previous studies to be a benign polymorphism [
4,
13,
22], even in the homozygous state [
13]. We found the R98Q substitution at similar frequencies in our cases and controls, further confirming that this polymorphism is unlikely to be disease producing.
Previous studies of
DJ-1 have often limited the scope of screening to detect only coding sequence changes. Only one third of the previously completed studies [
6,
12,
19,
21], which represent less than one quarter of the 1200 total patients screened to date, employed methods to detect any alternations in gene dosage. Therefore, it is possible that the frequency of
DJ-1 mutations might have been underestimated in these studies. However, our analyses would suggest that even when a cohort is enriched for families more likely to carry a
DJ-1 mutation and methods include screening for gene dosage alterations, mutations in this gene are rare in familial PD.
There are several notable strengths of our study. First, thorough screening of the
DJ-1 gene was performed, including sequencing of all translated exons as well as testing for genomic rearrangements by MLPA analysis. Second, families were selected for testing based on their evidence of linkage to the
DJ-1 region on chromosome 1p36. We have previously successfully utilized this approach to prioritize screening of our patient cohort for mutations at the
PRKN locus [
7]. Thus, we have augmented our tested sample for families most likely to have a mutation in
DJ-1. Conversely, our study also has several limitations. First, due to the likely paucity of
DJ-1 mutations, we did not screen our entire familial PD sample. However, we did screen more than 15% of our sample which included those most likely to harbor a
DJ-1 mutation. Second, we did not screen our entire sample of early onset PD cases, which is typically defined as having an age of onset of 45 or earlier. Rather we used a more cost effective approach to only screen families with evidence of linkage to this genomic region. Finally, the dosage methods did not interrogate either exon 2 or 4. In addition, positive controls carrying DJ-1 exon deletions were not available to us. While false negatives cannot be ruled out, MLPA analysis was performed at least twice for each sample minimizing this risk. However, based on the results of the screening in our sample as well as the results from other studies, mutations in
DJ-1 do not appear to be a common cause of PD, and are much rarer than mutations in the
PRKN or
LRRK2 genes.