5178

PEG3

PW1|ZKSCAN22|ZNF904|ZSCAN24;paternally expressed 3;PEG3;paternally expressed 3

Kruppel-type zinc finger protein|paternally-expressed gene 3 protein|zinc finger and SCAN domain-containing protein 24

19q13.4

protein-coding

OBJECTIVE: To measure mRNA expression levels of paternally expressed gene 3 (PEG3) in gynecologic cancer cell lines and to determine if DNA methylation is involved in the control of PEG3 expression. METHODS: PEG3 mRNA levels were measured with real-time PCR from 28 gynecologic cancer cell lines and compared to normal tissues. PEG3 mRNA expression was correlated to promoter methylation levels measured by real-time methylation-specific PCR. Polymorphism-specific restriction digestion was employed to analyze PEG3 allele distribution. RESULTS: While expressed in normal gynecologic tissues, PEG3 is silenced in all endometrial and cervical cancer cell lines studied. In the eight ovarian cancer cell lines, five were found to be PEG3 negative, the remaining three express low levels of PEG3 mRNA. In contrast, loss of maternal imprinting and relatively high PEG3 expression levels were detected in all four choriocarcinomas cell lines studied. No cell line confirmed to contain two copies of PEG3 expressed PEG3 mRNA, suggesting that PEG3 downregulation is not due to genetic deletion. PEG3 mRNA expression was, however, quantitatively correlated to its promoter methylation status. Treatment of PEG3 negative cells with DNA methyltransferase inhibitor 5-aza-deoxycytidine led to partial promoter demethylation and biallelic reactivation of PEG3 transcription, confirming the methylation-mediated mechanism for PEG3 silencing. CONCLUSION: PEG3 silencing is associated with DNA hypermethylation but not gene deletion in cell lines tested. These results suggest that loss of PEG3 expression may be a frequent event in gynecologic cancers. Given the known role of PEG3 in p53-mediated apoptosis, it is possible that PEG3 functions as a tumor suppressor.

BACKGROUND: Imprinted tumor suppressor genes may be particularly important in the pathogenesis of ovarian cancer. Two imprinted genes, paternally expressed 3 (PEG3) and aplasia Ras homologue member I (ARHI), are the most frequently down-regulated in ovarian cancers on gene expression arrays. METHODS: PEG3 and ARHI expression levels were evaluated with real-time reverse-transcriptase polymerase chain reaction (PCR) analysis. Promoter methylation was measured by pyrosequencing, and loss of heterozygosity (LOH) was detected by PCR-LOH assays. RESULTS: PEG3 was down-regulated in 75% and ARHI was down-regulated in 88% of 40 ovarian cancers. ARHI CpG islands I and II were hypermethylated in 13 of 42 ovarian cancers (31%) and in 5 of 42 ovarian cancers (12%), respectively, and hypermethylation was associated with reduced ARHI expression in all 18 samples of ovarian cancer with CpG island hypermethylation. PEG3 was hypermethylated in 11 of 42 ovarian cancers (26%), and PEG3 expression was down-regulated in 10 of those 11 cancers. LOH was detected in 8 of 35 informative cases for ARHI (23%) and in 5 of 25 informative cases for PEG3 (20%). PEG3 and ARHI expression was highly correlated in human ovarian cancers (correlation coefficient [R]=0.69; P< .0001). PEG3 and ARHI also were methylated concordantly in ovarian cancers (R=0.36; P= .019). Re-expression of PEG3, similar to that of ARHI, markedly inhibited ovarian cancer growth. ARHI and PEG3 expression could be restored by treatment with 5-aza-2-deoxycytidine and trichostatin A, consistent with the importance of promoter methylation and histone acetylation in regulating expression of both genes. CONCLUSIONS: Loss of expression of the growth-inhibitory imprinted genes ARHI and PEG3 through promoter methylation, LOH, and other mechanisms may stimulate clonogenic growth and contribute to the pathogenesis of a majority of ovarian cancers.