Epithelial and stromal interactions and the role of TGF-beta (TGF-b) in mammary development and breast cancer progression
Return to Lab Home Page
During the dry period between successive lactations the mammary gland of dairy cows undergoes extensive remodeling that is marked by phases of involution and mammogenesis. A dry period of 45 to 60 days between lactations is generally recommended to prevent milk production losses in a subsequent lactation (Coppock et al., 1974; Klein and Woodward, 1943; Swanson, 1965). Reducing the length of the dry period without loss of milk yield would improve efficiency of dairy production by reducing the number of unproductive days in the lifetime of a cow. Recently, renewed interest in the effects on postpartum health, reproduction, and milk production have provided conflicting evidence for the argument to shorten the dry period (Kuhn et al, 2006; de Feu et al., 20009; Watters et al., 2009). A better understanding of the mechanisms of mammary changes during the dry period may provide the evidence needed to develop a dry period strategy based on the biological needs of the cow. Changes in mammary epithelium during the dry period have been well characterized; however, few studies have examined the changes that occur in stromal tissue.
A. To characterize changes that occur in mammary stroma during the dry period mammary biopsies were taken from nine multi-gravid Holstein cows in late lact ation, at 1 week after dry-off, 3 weeks before expected calving date, and 1 week before expected calving date.
- Number of activated fibroblasts was greater at 1 week dry than 1 week before calving
- Percent intralobular stromal area was greater at 1 week dry (32%) and 3 weeks before calving (37%) than 1 week before calving (25%)
- Percent area of stromal TGF-b1 expression decreased 15% from late lactation to the dry period.
- The percentage of stromal area expressing fibronectin, MMP-3, and TGF-bR2 and the percentage of myofibroblasts were not different across biopsy stages.
These results support the concept that stromal expression of transforming growth factor-b1 and fibroblast proliferation may be important for remodeling during the dry period.
B. To identify differences in gene expression in mammary epithelial and intralobular stromal compartments of dry cows 3 weeks prepartum epithelial and intralobular stromal tissues were isolated with laser capture microdissection. Global gene expression was measured with Bovine Affymetrix GeneChips.
- Moderated t-tests from gene-specific linear model analysis with cell type as a fixed effect showed more than 3,000 genes were differentially expressed between tissues (P<0.05; FDR<0.17).
- Epithelial signatures were enriched with gene sets for protein synthesis, metabolism and secretion.
- Stromal signatures were enriched with genes that encoded molecules important to signaling, extracellular matrix composition and remodeling.
- Transcriptome differences also showed evidence for paracrine interactions between tissues in stimulation of IGF1 signaling pathway, stromal reaction, angiogenesis, neurogenesis, and immune response.
Molecular signatures point to the dynamic role the stroma plays in prepartum mammogenesis and highlight the importance of examining the roles of cell types within the mammary gland when targeting therapies and studying mechanisms that affect milk production.
C. To determine the effect of exogenous TGF-b on changes in the mammary gland during the dry period, biopsied tissue was cut into explants and incubated with TGF-b1 for 2 hrs. We found that treating mammary tissue collected from cows at 1 week dry with TGF-b1 increased epithelial and stromal cell proliferation as well as activation of stromal fibroblasts into myofibroblasts. TGF-b1 treatment may be an effective way to shorten the dry period by hastening mammary remodeling and thus impacting milk production and increase the profitability of dairy farming.
Understanding the role of stroma in breast cancer progression
Although breast cancer is an epithelial cell phenomenon, our previous work and work of others suggest that the surrounding stroma acts as a mediator of tumor invasion and metastasis. Global gene expression analysis of the differential contribution of tumor epithelium and stroma to the invasive breast cancer transcriptome, suggested that invasion occurs through the acquisition of a motile phenotype in tumor epithelial cells and a reactive phenotype in cancer associated fibroblasts [CAFs]. Malignant epithelial cells acquire the molecular signature of a motile phenotype prior to invasion and reactive stroma likely forms in response to malignant transformation of epithelial cells.
Breast tumor cells gain motility through multiple mechanisms including reorganization of the actin cytoskeleton, regulation of focal adhesion, and epithelial to mesenchymal transition. The down regulation of mRNA expression of basement membrane proteins in tumor epithelial cells and up-regulation of extracellular matrix proteases in CAFs also likely contributes to the tumor epithelial cell ability to transverse the basement membrane barrier during invasion. Further ECM dense reactive stroma associated with epithelial tumors, has been proposed to serve as a scaffolding for cancerous epithelial cells to use during invasion as well as structural support for angiogenesis.
Transforming growth factor betas (TGFB) are multifunctional cytokines which inhibit epithelial cell growth, stimulate epithelial to mesenchyme transition, regulate ECM deposition and degradation, and modulate immune function and wound repair. These multiple functions attributed to TGFB1 in normal tissue lead to paradoxical roles in breast cancer development. TGFB1 has been proposed to both suppress tumor growth as well as promote breast tumor invasion and metastasis. Studies using mouse models of breast cancer suggest that TGFB1 switches its role from a tumor repressor in early stage mammary tumors to a tumor promoter in late stage tumors, and may also increase the risk of metastases. Previously, we showed that higher expression of TGFBR2 in breast tumor stroma was correlated with a poor prognosis, and that patients with recurrent breast cancer had higher expression of TGFB1 in the epithelial component of primary tumors. Differences in global gene expression between epithelial and stromal tissues and normal breast tissue and invasive tumors led us to hypothesize that changes in gene expression leading to or enhancing the motile phenotype of tumor epithelium may be stimulated in part by the up-regulation of TGFB1 in CAFs. We proposed that TGFB1 facilitates breast cancer invasion by stimulating secretion of ECM proteins and proteases from CAFs, which results in remodeling of the ECM and creates an environment that promotes invasion.
Previous studies designed by us to test this hypothesis measured differences in fibroblasts and CAFs in vitro and the effects of TGFB1 treatment on primary fibroblast cultures. We found that CAFs were measurably different from normal fibroblasts in their response to TGFB1. Percent myofibroblasts expressing SMA was not different between primary fibroblasts isolated from normal breast tissue and invasive breast cancer. There was also no difference in rate of MDA-MB 231 in vitro invasion towards condition media collected from normal fibroblasts versus CAF primary cultures. In contrast, TGFB1 treatment significantly increased the percent of myofibroblasts and the invasion rate was greater in chambers with conditioned media from TGFB1 treated CAF versus normal fibroblast cultures treated with TGFB1. TGFB1 treatment significantly increased expression of fibronectin and laminin in CAF cultures, but did not have a significant effect in normal fibroblast cultures.
Using immunohistochemistry we measured expression of TGFB1 and its type II receptor (TGFBR2) in tumor stroma and epithelium in the same patient population used in our previous studies. Our objective was to determine if there was a relationship between protein expression of either TGFB1or TGFBR2 or both and changes in vimentin (VIM), smooth muscle alpha actin (SMA), fibronectin (FN) and laminin expression in invasive breast tumors. The relationship between tissue specific expression of these proteins and prognostic factors (i.e. estrogen receptor-ER, progesterone receptor PR, Erb2 and lymph node status as well as stage and tumor size) was analyzed by regression analysis. We report that mean expression of TGFB1 was significantly lower in both stromal and epithelial components of invasive tumors compared to normal breast tissue. Percent area of stromal VIM, SMA and FN was significantly greater in invasive breast cancer than normal tissue. ER negative tumors had higher epithelial VIM expression and higher expression of SMA in stroma compared to ER positive tumors. Stromal expression of VIM, SMA, FN and laminin were positively correlated with TGFBR2 in invasive breast cancer, but showed no significant relationship in normal breast tissue.
These studies (bovine mammary development and breast cancer) suggest that TGF-b, fibroblast activation, and epithelial-mesenchyme transdifferentiation play a pivotal role in bovine mammary development and neoplastic progression of human breast cancer. We are currently analyzing our data to design studies targeted at elucidating roles and mechanisms of these processes, with the aim to develop targeted therapies for hastening the dry period in dairy cows and preventing invasion and metastasis of breast cancer.