Oestrogen-dependent postmenopausal breast cancer is associated with the increased
expression of aromatase in breast adipose stromal cells (ASCs). Androgens are converted into
oestrogens by aromatase and tumour-derived factors such as prostaglandin E2 (PGE2)
stimulate aromatase expression via the activation of the proximal promoter PII. As a tumour
suppressor, p53 is often mutated in breast cancer. However, mutations in p53 in ASCs are
infrequent. This study aimed to determine the effect of PGE2 on p53 and to examine the role
of p53 in regulating aromatase expression in human breast ASCs in the context of obesity related postmenopausal breast cancer.
Primary human breast ASCs were isolated from breast tissue following breast reduction
surgery. ASCs were treated with 1μM PGE₂ or the PGE₂ mimetic forskolin (FSK;
25μM)/phorbol ester (PMA; 4nM), and/or RITA (1μM; to stabilize p53). The effect of PGE₂
on p53 expression, subcellular localization and activity was examined by real-time PCR,
Western blotting and immunofluorescence (IF). Results demonstrate that p53 transcript and
nuclear protein expression are significantly decreased in ASCs in response to PGE₂.
Moreover, PGE₂ also reduces p53 phosphorylation at serine 15 (Ser15) and inhibits its
nuclear localisation in ASCs. The effect of p53 on aromatase transcript and activity was
examined in RITA-treated ASCs by real-time PCR, and tritiated-water release assays. RITA
significantly decreased the PGE2-mediated mRNA and protein expression and activity of
aromatase. To examine whether p53 binds to PII directly, ChIP and reporter assays were also
performed to examine p53 binding and regulation of aromatase PII. ChIP results demonstrate
that p53 binds to a region on PII 438-418bp upstream of the transcription start site and that
this interaction is significantly decreased in the presence of PGE2. QPCR and reporter assays
demonstrate that RITA also significantly decreases the PGE2-stimulated activity of aromatase
PII expression and activity. IF performed on clinical samples of breast tissue from cancer-freewomen and women with breast cancer demonstrates that nuclear p53 expression is
significantly decreased in the latter group. We also found that the ratio of perinuclear to
nuclear p53 is increased in tumour-associated ASCs and a positive correlation between
perinuclear (inactive) p53 and aromatase was also observed.
In conclusion, our findings demonstrate that p53 inhibits aromatase expression via directly
binding to PII, and that this inhibition is alleviated in the presence of PGE₂. This study
therefore reveals an unconventional role for p53 as a tumour suppressor in breast cancer.