Single factors alone can induce mesenchymal-like morphology, but not promote full EMT in breast cancer cell lines with different hormone statuses.

BACKGROUND: Epithelial to mesenchymal transition (EMT) is considered to be important for cancer invasion and metastasis. Tumour hypoxia, in addition to Transforming Growth Factor-ß (TGF-ß) and Notch, amongst others, have been suggested to be involved in EMT. We therefore investigated if hypoxia, TGF-ß1 and the Notch ligand Jagged-1 alone induced morphological changes with corresponding EMT signatures in different epithelial breast cancer cell lines in vitro. Furthermore, we also studied whether or not TGF-ß1, or Jagged-1 in combination with hypoxia added any effect on EMT. METHODS: The cells were exposed to normoxia or hypoxia alone or in combination with TGF-ß1 or Jagged-1. Morphological responses to treatment were investigated by light microscopy, and changes in markers for EMT and hypoxia were evaluated by western blot analysis and immunofluorescence studies. RESULTS: One of the four cell lines (MCF7) became elongated and highly multipolar, indicative of EMT, following hypoxia, TGF-ß1 and Jagged-1 treatment per se with the most distinct morphological shift seen with Jagged-1 treatment in combination with hypoxia. Also, when regarding hypoxia, MCF7 cells showed the greatest change in EMT-markers of the four cell lines tested, but these changes were not consistent with a typical EMT pattern. The morphology of BT474 cells was not altered following Jagged-1 treatment, however, Jagged-1 increased E-cadherin levels. Morphology was changed following TGF-ß1 treatment of BT474 cells, but it did not affect E-cadherin levels. Neither Jagged-1 nor TGF-ß1 altered the levels of Vimentin in the BT474 cell line. The E-cadherin responses to hypoxia varied with end-point in both MCF7 and BT474 cells, and in most cases were not consistent with EMT. CONCLUSION: Our results using four different breast cancer cell lines in vitro do not provide evidence that EMT is induced by hypoxia alone or in combination with TGF-ß1 or the Notch ligand Jagged-1. The inconsistency in morphological appearance and EMT-markers, as well as the time dependent variation in E-cadherin responses could not support EMT. Importantly, there was not one single common response pattern to the stimuli used, suggesting that cell lines with different hormone statuses display individual traits that respond differently to the stimuli applied. Thus, based on the present results, common statements that single factors by themselves can induce EMT seem questionable.

Stromal Integrin α11ß1 Affects RM11 Prostate and 4T1 Breast Xenograft Tumors Differently.

PURPOSE: It has been implied that the collagen binding integrin α11ß1 plays a role in carcinogenesis. As still relatively little is known about how the stromal integrin α11ß1 affects different aspects of tumor development, we wanted to examine the direct effects on primary tumor growth, fibrosis, tumor interstitial fluid pressure (PIF) and metastasis in murine 4T1 mammary and RM11 prostate tumors, using an in vivo SCID integrin α11-deficient mouse model. METHODS: Tumor growth was measured using a caliper, PIF by the wick-in-needle technique, activated fibroblasts by α-SMA immunofluorescence staining and fibrosis by transmission electron microscopy and picrosirius-red staining. Metastases were evaluated using hematoxylin and eosin stained sections. RESULTS: RM11 tumor growth was significantly reduced in the SCID integrin α11-deficient (α11-KO) compared to in SCID integrin α11 wild type (WT) mice, whereas there was no similar effect in the 4T1 tumor model. The 4T1 model demonstrated an alteration in collagen fibril diameter in the integrin α11-KO mice compared to WT, which was not found in the RM11 model. There were no significant differences in the amount of activated fibroblasts, total collagen content, collagen organization or PIF in the tumors in integrin α11-deficient mice compared to WT mice. There was also no difference in lung metastases between the two groups. CONCLUSION: Deficiency of stromal integrin α11ß1 showed different effects on tumor growth and collagen fibril diameter depending on tumor type, but no effect on tumor PIF or development of lung metastasis.

The Effect of Stromal Integrin ß3-Deficiency on Two Different Tumors in Mice.

There is an increasing focus on the tumor microenvironment in carcinogenesis. Integrins are important receptors and adhesion molecules in this environment and have been shown to be involved in cell adhesion, proliferation, differentiation and migration. The present study aimed to evaluate the effect of stromal integrin ß3-deficiency on tumor growth, angiogenesis, interstitial fluid pressure (PIF), fibrosis and metastasis in a murine breast cancer (4T1) and a prostate tumor (RM11) model. We showed that stromal integrin ß3-deficiency led to an elevation in PIF that correlated to a shift towards thicker collagen fibrils in the 4T1 mammary tumor. In the RM11 prostate carcinoma model there was no effect of integrin ß3-deficiency on PIF and collagen fibril thickness. These findings support the notion that changes in the collagen scaffold influence PIF, and also indicate that there must be important crosstalk between the stroma and tumor cells, in a tumor cell line specific manner. Furthermore, stromal integrin ß3-deficiency had no effect on tumor growth or angiogenesis in both tumor models and no effect on lung metastasis in the 4T1 mammary tumor model. In conclusion, the stromal ß3 integrin influence PIF, possibly via its effect on the structure of the collagen network, in a tumor cell line dependent manner.