It is estimated that more than 42,000 new cases and 7,780 deaths will be attributed to endometrial cancer (EC) during 2009. Considering the anticipated number of deaths from EC has nearly tripled since 1987 (2,900), a mandate exists to develop new paradigms focusing on prevention, early detection and treatment. Detailed cataloging of the phenotypic, biochemical and genetic signatures of EC will facilitate the maturation of a risk profile for EC thereby affording patients at risk the opportunity for enrollment in prevention or early therapeutic trials.

Endometrioid carcinoma (or type I EC) accounts for approximately 80% of endometrial cancers. Epidemiological studies have provided convincing evidence for the pivotal role of obesity in the etiology of this disease. Adiposity induced hyperinsulinemia and hyperestrogenemia are compounding factors and result in physiological alterations that collectively have been reported to facilitate mitogenesis, stimulate neoplastic growth and inhibit apoptosis. The stimulatory effects of estrogen on the endometrium are well characterized. Furthermore, there is increasing evidence that hyperinsulinemia is more directly involved in carcinogenesis. Insulin functions as a growth factor by increasing endometrial insulin receptors thereby enhancing the growth potential of the endometrium to IGFs which are up regulated by estrogens and considered strong mitogens that stimulate cell proliferation and suppress apoptosis. Insulin also inhibits serum and tissue levels of IGFBP thereby increasing the bioavailability of IGF. While the above metabolic abnormalities facilitate incessant proliferative, the evolutionary mechanisms initiating cell transformation from proliferative endometrium to CAH (complex atypical hyperplasia), the precursor lesion, and EC remain unknown. The potential mechanisms include mutations, gene silencing or inactivation of chromatin or a combination of these processes. Microsatellite instability (MSI), reflecting silencing of one or more mismatch repair genes, has been witnessed in approximately 30% of endometrioid EC and 70% of the MSI cohort correlated with methylation of MLH1. Multiple additional major genetic alterations including inactivation of tumor suppressor genes and mutations in oncogenes have been reported. Mutations in PTEN, which antagonizes the PI3K/AKT pathway, have been reported in as high as 83% of EC and 55% of precursor lesions. PTEN mutations are more frequently associated with MSI+ (60–80%) than with MSI‐ (24–35%) lesions and have a favorable prognosis; patients with PTEN silencing due to promoter methylation have a poorer prognosis. PIK3CA mutations occurring within the catalytic subunit of PI3K constitutively activate AKT and have been detected in 39% of EC but only in 7% of CAH suggesting a later event than PTEN mutations. Mutations in K‐ras and β‐catenin have been observed in 10–35% of EC resulting in constitutive activation of the MAPK and β‐catenin pathways, respectively. The observations that 42–69% of K‐ras mutations occur in MSI+ EC and that the majority are G to A transition mutations suggest the silencing of MGMT as a possible etiology.

Gene silencing via methylation of gene promoters is a common phenomenon witnessed in nearly all solid tumors. Methylation of cytosine in CpG islands results in the binding domain proteins recruiting histone deacetylases. The subsequent deacetylation and methylation of histone lysine result in tight histone/DNA interactions which impede transcription. While DNMT1 is necessary for maintenance methylation, DNMT3A and 3B are considered responsible for the de novo methylation at the 5‐position of cytosine. Several genes have been reported to be silenced via promoter methylation in EC including MLH1, APC (critical component in the β‐catenin pathway), PTEN (as noted above), estrogen receptors and progesterone receptors. Several additional cogent tumor suppressor genes or receptors silenced in other hormone responsive tissues such as ovary, breast and prostate and functional in the four major molecular pathways functioning in the endometrium are candidates for promoter CpG methylation. These include MGMT, IGFBP3, RUNX3, and CDKN2A.

We have reported varying levels of DNMT mRNA and protein expression in EC tissue specimens and cell culture; DNMT1 and DNMT3B were up regulated in tissues and cell culture while DNMT2 and 3A did not differ significantly from controls. The addition of estrogen to Ishikawa cells (estrogen receptor positive) resulted in an up regulation of DNMT3B. Considering progesterone receptors are absent in many moderate and essentially all poorly differentiated EC, PRB promoter methylation was assessed in Hec‐1B and KLE cells before and after treatment with methyltransferase (MT) and histone deacetylase (HDAC) inhibitors alone or in combination. The PRB promoters in both cell lines were methylated prior to treatment with ADC (aza‐deoxycytidine), a methyltransferase inhibitor, and became demethylated following exposure to ADC. PR‐B mRNA expression in both HEC‐1B and KLE cells increased in response to both MT and HDAC inhibitors with a synergistic effect noted with the combination of ADC and TSA. Western blot confirmed the corresponding PR‐B protein expression levels. The effect of TSA on DNMT mRNA expression was analyzed in Ishikawa cells. TSA did not significantly alter DNMT1 or ‐3A mRNA levels but markedly suppressed DNMT3B mRNA expression. RNA decay curves demonstrated rapid degradation of DMNT3B mRNA following TSA treatment of both Ishikawa and AN3 cells. Recent reports have demonstrated a direct interaction of miRNA 29b with the 3′ untranslated regions of both DNMT3A and DNMT3B in lung cancer and acute myeloid leukemia resulting in downregulation of both DNMT3A and DNMT3B.

Further studies of the epigenome in endometrial cancer and other solid tumors are required to test the utility of using methylation markers for diagnostic and prognostic purposes and facilitating strategies in the development of epigenetic therapies.

Citation Information: Cancer Prev Res 2010;3(1 Suppl):CN02-02.