The relationship between reproductive history and breast cancer is complex. A full term pregnancy in younger women is protective against breast cancer later in life, while a first full term pregnancy after 40 years of age is associated with higher cancer occurrence. In addition, there is an increase in breast cancer within the first two years of pregnancy, regardless of age. Pregnancy associated breast cancer also presents at a later clinical stage and is more invasive than non-pregnancy associated breast cancers. We hypothesize that the cancer promotional effect of pregnancy is due to physiologic remodeling of the mammary stroma that occurs after pregnancy. During parturition or weaning-induced mammary gland involution, mammary stroma is degraded and stroma levels of matrix metalloproteinases (MMP) and pro-inflammatory mediators are elevated. We propose that these involution-associated changes in mammary stroma promote an epithelial to mesenchymal transition (EMT) in pre-existing neoplastic cells. The resulting activated tumor cells, coupled with the weakened stromal barrier, increases the likelihood of tumor cell extravasation. Since the likelihood of having occult tumor cells increases with age, pregnancy associated gland involution is more likely to effect EMT and metastases in older women.
In Aim 1. the hypothesis that tumor cell metastatic potential is modulated by reproductive status will be tested in vivo using xenograft models. Human mammary tumor cells will be injected directly into teats or mammary fat pads of immunodeficient mice at different reproductive stages (i.e. nulliparous, 'pregnancy-associated', and parous). The ability of tumor cells to exit the gland is predicted to be dependent on the degree of tissue remodeling at the time of injection: low in nulliparous and parous, and high in regressing glands. Primary tumors will be evaluated histologically and metastases to lung, liver, brain and long bone of the leg will be measured by semi-quantitative PCR using a unique human sequence, the CCR5 gene.
In Aim 2, the hypothesis that endocrine-induced changes in stroma composition can either promote or suppress EMT will be evaluated in vitro using reconstitution models. Human mammary epithelial cells, which differ in metastatic potential from non-tumorigenic to metastatic, will be combined with mammary stroma isolated from rats with different reproductive states. It is anticipated that quiescent mammary stroma (nulliparous and parous) will inhibit, while involuting matrix will promote EMT in highly metastatic cells. Functional evidence for EMT will be obtained by measuring stroma-dependent changes in cell adhesion, motility and invasiveness using modified Boyden chambers assays. Morphological and biochemical evidence for EMT will be evaluated in a 3-dimensional model of tumor formation. The effects of mammary matrix on key markers of EMT, actin fiber organization, and expression of E-cadherin, vimentin, gelsolin and Fspl proteins, will be determined biochemically.
In Aim 3, the hypothesis that involuting mammary matrix induces, while nulliparous and parous mammary matrices suppress expression of MMP in tumor cells will be evaluated by reporter gene assays. Luciferase-based reporter constructs containing either the stromelysin (MMP-3), gelatinase A (MMP-2), or gelatinase B (MMP-9) promoters will be tested for activity in tumor cells plated on different mammary matrices. The results of these studies will contribute to our understanding of the effects pregnancy-associated changes in mammary stroma have on breast cancer progression, and should provide insights into epidemiological data describing the dual effect of pregnancy on breast cancer.
