INTRODUCTION Mdm2 inhibits p53 transactivation by forming a p53-Mdm2 complex on chromatin. Upon DNA damage-induced complex disruption, such latent p53 can be activated, but in cells overexpressing Mdm2 because of a homozygous single nucleotide polymorphism at position 309 (T --> G) of mdm2, the complex is highly stable and cannot be disrupted by DNA damage, rendering p53 inactive. METHODS To determine whether the p53 response phenotype is influenced differentially in cells with variable mdm2 genotypes, we compared responses to DNA damage and targeted p53-Mdm2 complex disruption by Nutlin-3 in the following wild-type p53 human cancer cell lines: A875 and CCF-STTG-1 (G/G for mdm2 SNP309), SJSA-1 (mdm2 genomic amplification and T/T for mdm2 SNP309), MCF-7 (estrogen-induced Mdm2 overexpression and T/G for mdm2 SNP309), ML-1 and H460 (T/T for mdm2 SNP309), and K562 (p53-null and T/G for mdm2 SNP309). We also examined mdm2 gene-splicing patterns in these lines by cloning and sequencing analyses. RESULTS While Mdm2-overexpressing G/G cells were resistant to p53 activation by DNA damage, they were sensitive to Nutlin-3. Strikingly, the p53 G1 checkpoint in G/G cells was activated by Nutlin-3 but not by etoposide, whereas in other Mdm2-overexpressing cells, both drugs activated p53 and subsequent G1 arrest or apoptosis. cDNA clones lacking exons 5-9 were generated at a high frequency in cells overexpressing Mdm2. CONCLUSION Nutlin-3 and DNA damage distinguish a differential phenotype in human cancer cells with G/G mdm2 SNP309 from other Mdm2 overexpressers. Categorization of the Mdm2 isoforms produced and their influence on p53 activity will help in characterization and treatment development for different cancers.

Due to the established association between estrogen levels and breast cancer risk, polymorphic variation in genes regulating estrogen levels is thought to be related to breast cancer risk. Aromatase, the protein product of the CYP19 gene, is involved in the production of endogenous estrogens via androgen conversion. We examined whether polymorphic variation in CYP19 associated with increased breast cancer risk in a population based case-control study. We examined two single nucleotide polymorphisms (SNP), rs1008805 (A/G) and rs730154 (C/T), which have been shown to tag SNPs within two different haplotype blocks in CYP19. Among premenopausal women, the presence of at least one G allele at rs1008805 was significantly associated with an increase in the risk of breast cancer (OR = 1.72 [95% CI, 1.20-2.49]), especially with estrogen and progesterone receptor negative breast cancer (OR = 3.89 [1.74-8.70] and OR = 2.52 [1.26-5.05], respectively). No association was observed among postmenopausal women (OR = 1.06 [0.82-1.36]). There was no significant association between rs730154 and breast cancer, regardless of menopausal status. Our results suggest that premenopausal women carrying the G allele at CYP19 rs1008805 have increased risk of breast cancer. The finding supports the potential role of variation in estrogen biosynthesis genes in premenopausal breast cancer risk.

10-Decarbamoyl-mitomycin C (DMC), a mitomycin C (MC) derivative, generates an array of DNA monoadducts and interstrand cross-links stereoisomeric to those that are generated by MC. DMC was previously shown in our laboratory to exceed the cytotoxicity of MC in a human leukemia cell line that lacks a functional p53 pathway (K562). However, the molecular signal transduction pathway activated by DMCDNA adducts has not been investigated. In this study, we have compared molecular targets associated with signaling pathways activated by DMC and MC in several human cancer cell lines. In cell lines lacking wild-type p53, DMC was reproducibly more cytotoxic than MC, but it generated barely detectable signal transduction markers associated with apoptotic death. Strikingly, DMCs increased cytotoxicity was not associated with an increase in DNA double-strand breaks but was associated with early poly(ADP-ribose) polymerase (PARP) activation and Chk1 kinase depletion. Alkylating agents can induce increased PARP activity associated with programmed necrosis, and the biological activity of DMC in p53-null cell lines fits this paradigm. In cell lines with a functional p53 pathway, both MC and DMC induced apoptosis. In the presence of p53, both MC and DMC activate procaspases; however, the spectrum of procaspases involved differs for the two drugs, as does induction of p73. These studies suggest that in the absence of p53, signaling to molecular targets in cell death can shift in response to different DNA adduct structures to induce non-apoptotic cell death.