Estrogen Receptor α (ERα) is a key regulator in normal mammary gland development and acts as a growth factor in breast cancer development. We have developed the CERM mouse model with targeted ERα overexpression and deregulation specifically to the mammary gland. By 4 months of age, the CERM mice develop ERα-initiated mammary ductal hyperplasia and ductal carcinoma in-situ which mimic human disease (Frech et al., 2005). In women, there is evidence that early full-term pregnancy decreases lifetime ERα-positive breast cancer risk. Paradoxically, there is also evidence that breast cancer risk is transiently increased within ten years of early full-term pregnancy (Schedin, 2006). One potential mechanism for the lifetime protective effect is believed to occur via differentiation (Russo et al., 2006). In addition, it is well established that STAT5a, a latent cytoplasmic transcription factor, is highly activated during pregnancy and lactation, and is involved in alveolar proliferation and terminal differentiation (Liu et al, 1996 & 1997). Therefore, we hypothesized that Stat5a contributes to the protective effects of pregnancy. The CERM mice develop hyperplastic alveolar nodules (HANs) by 12 months of age, therefore we postulated that pregnancy would protect against HANs in the CERM mice, which have Stat5a, and that loss of Stat5a alleles would reduce the protective effects of pregnancy.


The CERM mouse model, which has targeted ERα transgene expression to mammary epithelial cells, was bred with the Stat5a knockout mouse model to generate CERM, CERM/Stat5a+/- and CERM/Stat5a-/- mice. The mice went through several pregnancies and lactation cycles before being necropsied at 12-13 months of age.


At 12 months of age, CERM mice showed an increased incidence of mammary hyperplastic alveolar nodules (HANs) (p= 0.0254) over wild type (WT) mice. That increase is reduced with loss of one Stat5a alleles (CERM/Stat5a+/-) and further reduced to zero (p=0.0022) with loss of 2 Stat5a alleles (CERM/Stat5a-/-). We observed an unexpected finding with parity in the CERM and CERM/Stat5a+/- mice. At 12 months of age, the incidence of HANs was not reduced in multiparous CERM versus nulliparous CERM mice as well as in multiparous CERM/Stat5a+/- versus nulliparous CERM/Stat5a+/- mice. Since the incidence of HANs was zero in the nulliparous CERM/Stat5a-/- mice, the protective effects of pregnancy were not studied with loss of two Stat5a alleles. However, we had generated one 12 month old parous CERM/Stat5a-/- mice, which interestingly developed HANs as opposed to nulliparous CERM/Stat5a-/- mice, which do not develop HANs. Interestingly, additional preliminary data demonstrated that parity in the CERM and CERM/Stat5a+/- mice can lead to the development of different neoplastic lesions that were variably ERα positive or negative.


Pregnancy in our CERM model did not confer protection against HANs development at 12 months of age. This lack of pregnancy protection seemed to occur via a Stat5a-independent ERα-dependent growth pathway. Since Stat5a loss leads to lack of HAN development in nulliparous mice, Stat5a loss seemed to have a different impact in multiparous mice which showed development of HANs. Further studies will determine which signaling pathways are involved in lack of pregnancy protection in our models.

Citation Information: Cancer Prev Res 2008;1(7 Suppl):A6.

Seventh AACR International Conference on Frontiers in Cancer Prevention Research-- Nov 16-19, 2008; Washington, DC