Credit: CORBIS

There is increasing evidence to suggest that the Parkinson's disease susceptibility genes are also associated with cancer. Indeed, patients with Parkinson's disease seem to have an increased risk of developing melanoma. Mark Arends and colleagues now show that PARK2 , the loss of which causes autosomal recessive juvenile Parkinson's disease, is a haploinsufficient suppressor of colorectal cancer (CRC).

There is increasing evidence to suggest that the Parkinson's disease susceptibility genes are also associated with cancer.

Large amplifications and deletions on specific chromosomes are one characteristic of CRC. Arends and colleagues investigated whether smaller chromosomal alterations might also be associated with CRC and could be used to identify susceptibility genes. Analyses of more than 100 primary CRC samples and previously published data sets revealed that small copy number losses were common to chromosome 6q26, which encodes PARK2. They also found small deletions of the PARK2 promoter as well as others that also covered the adjacent PARK2 co-regulated ( PACRG ) gene. In total, 33 of 100 sporadic CRC samples had copy number losses affecting PARK2: most of these were heterozygous deletions, point mutations were rarely found and there was some evidence of hypermethylation of the PARK2 promoter. Real-time quantitative PCR revealed that PARK2 transcript levels were reduced in tumour samples with allelic losses at chromosome 6q26, and this correlated with reduced PARKIN (the protein product of PARK2) expression.

So, is PARK2 loss important for CRC development? The authors analysed published transcriptional profiles from microdissected CRC samples and found that the reduced expression of adenomatous polyposis coli ( APC ) — a tumour suppressor commonly mutated in CRC — correlated with reduced PARK2 expression, indicating that the expression of both genes is coordinately downregulated in CRC. The authors found that overexpression of PARKIN in CRC cell lines reduced proliferation and that this was significantly more pronounced in APC-deficient cell lines. Next, the authors studied Park2+/− mice and found that they did not develop intestinal adenomas; however, crossing with ApcMin/+ mice (which do develop adenomas) increased the prevalence of adenomas fourfold compared with ApcMin/+ mice. The remaining wild-type Park2 allele was not lost in the adenomas that formed, suggesting that Park2 is a haploinsufficient tumour suppressor.

PARK2 loss has also been observed in ovarian, breast and lung carcinomas, and glioblastoma, indicating that its role as a tumour suppressor might not be specific to colon epithelium. It will be intriguing to identify the underlying mechanisms that connect PARKIN, an E3 ubiquitin ligase, to tumour suppression.