Are there any more ovarian tumor suppressor genes? A new perspective using ultra high-resolution copy number and loss of heterozygosity analysis

• Published in: Genes chromosomes & cancer 2009-10, Vol.48 (10), p.931-942
• Main Authors: Gorringe, Kylie L., Ramakrishna, Manasa, Williams, Louise H., Sridhar, Anita, Boyle, Samantha E., Bearfoot, Jennifer L., Li, Jason, Anglesio, Michael S., Campbell, Ian G.
• Format: Article
• Language: English
• Subjects: Chromosomes, Human, Pair 15; Chromosomes, Human, Pair 8; Computational Biology - methods; Female; Gene Deletion; Gene Dosage; Gene Expression Profiling; Genes, Tumor Suppressor; Humans; Loss of Heterozygosity; Oligonucleotide Array Sequence Analysis; Ovarian Neoplasms - classification; Ovarian Neoplasms - genetics; Polymorphism, Single Nucleotide; Statistics, Nonparametric
• ISSN: 1045-2257; 1098-2264
• DOI: 10.1002/gcc.20694

Ovarian cancer is characterized by complex genetic alterations, including copy number loss and copy number‐neutral loss of heterozygosity (LOH). These alterations are assumed to represent the “second hit” of the underlying tumor suppressor gene (TSG), however, relative to the number of LOH hotspots reported, few ovarian TSGs have been identified. We conducted a high‐resolution LOH analysis using SNP arrays (500K and SNP6.0) of 106 primary ovarian tumors of various histological subtypes together with matching normal DNA. LOH was detected in at least 35% of samples on chromosomes 17, 19p, 22q, Xp, 13q, 8p, 6q, 4q, 5q, 1p, 16q, and 9q with a median minimal region of overlap of only 300 kb. Subtype‐specific differences in LOH frequency were noted, particularly for mucinous cases. We also identified 192 somatic homozygous deletions (HDs). Recurrent HDs targeted known TSGs such as CDKN2A (eight samples), RB1 (five samples), and PTEN (three samples). Additional recurrent HDs targeted 16 candidate TSGs near minimal regions of LOH on chromosomes 17, 13, 8p, 5q, and X. Given the importance of HDs in inactivating known genes, these candidates are highly likely to be ovarian TSGs. Our data suggest that the poor success of previous LOH studies was due to the inability of previous technology to resolve complex genomic alterations and distinguish true LOH from allelic imbalance. This study shows that recurrent regions of LOH and HD frequently align with known TSGs suggesting that LOH analysis remains a valid approach to discovering new candidates. © 2009 Wiley‐Liss, Inc.