Matrix density-induced mechanoregulation of breast cell phenotype, signaling and gene expression through a FAK-ERK linkage.

PP Provenzano, DR Inman1, KW Eliceiri, and PJ Keely
1Department of Pharmacology, University of Wisconsin, Madison, WI, USA; 2Department of Biomedical Engineering, University of Wisconsin, Madison, WI, USA; 3Laboratory for Optical and Computational Instrumentation, University of Wisconsin, Madison, WI, USA and 4University of Wisconsin Paul P. Carbone Comprehensive Cancer Center, University of Wisconsin, Madison, WI, USA

Claiming the spot as the second most common type of cancer, breast cancer is a prevalent problem for women all over the world. As a result, many significant studies have been carried out and have yielded a number of risk factors for the disease. In this paper by Provenzano et al. from the University of Wisconsin, the authors sought to explore the mechanism by which these purported causes lead to breast cancer through the mechanoregulation of cell phenotype. In particular, Provenzano et al. delves into the signaling cascades that arise from mammographically dense breast tissue, one of the best indicators for the onset of developing breast carcinoma. Previous research has shown that regions of high tissue density in breasts are associated with high stromal collagen and cell content. The authors set out in a logical manner to test whether or not increased collagen content alone (without stromal cells) resulted in cancerous/invasive phenotypes in non-transformed epithelial cells.

Fig. 1

First, the authors examined the effects of varying matrix (collagen) density on tubule structures in non-transformed mammary epithelial cells (MECs). From mechanical testing (described in a referenced paper) of the matrix, it was found that higher matrix density means an increased matrix stiffness. Byusing this knowledge, the authors cultured MECs under both low density (optimal) and high density conditions. A comparison of the cell morphology revealed normal tubule ducts in low density matrices and unstable cell-cell adhesions in high density matrices. Invasive phenotypes were also seen in cells lining the epithelial-high density ECM border. Further changes in boundary (matrix) conditions to influence collagen contraction in the matrix gel revealed that force balance at the cell-matrix interface and the physical properties of the immediate cell environment affect the phenotype of epithelial cells surrounding the breast tissue.

Cells mechanically interact with their environment mainly through the focal adhesion (FA). Knowing that these integrin-containing focal adhesions actively regulate epithelial phenotype, the authors observed and quantified the state of FAK, a biochemical signal that arises from FA mechanosensing, versus the matrix modulus. Their results showed that stiffer matrices were associated with increased phosphorylated FAK (Western Blotting). In addition, immunofluorescent microscopy and measurement of the cell perimeter confirmed more invasive phenotypes in stiffer matrices. It was also shown that FAK is force-responsive in MECs; mechanical stretching of collagen-coated 2D substrates and 3D collagen matrices with seeded MECs led to increased levels of phosphorylated FAK in both transformed and non-transformed cells. This further supports their claim that force balance at the cell-matrix boundary is important in regulating cell phenotype.

Fig. 2

To elucidate the importance of FAK in influencing downstream cancerous behavior in cells, Provenzano et al. took a more biochemical approach. By using RNA interference, the research group knocked out FAK in high density collagenous cultures to show a reversion of cellular phenotype from protrusions and dense actin stress-fibers to a lessening of both. Inhibition of biochemical signals downstream of FAK had similar effects. Through selective inhibition of the downstream Rho/ROCK pathway after the formation of invasive phenotype in the cells, the authors observed a reversion of cellular characteristics as previously mentioned for FAK inhibition. This suggests that both FAK and Rho are part of the mechanism used by MECs in their response to mechanical stimuli.

Fig. 3

Building upon the FAK and Rho/ROCK pathways, the authors also looked at the ERK/MAPK pathway, an implicated pathway in the cellular proliferation, migration, and apoptosis. By using the same biochemical approach as before, the authors were able to show that ERK activation is actually regulated by FAK and that the inhibition of ERK led to the suppression of MEC proliferation. Principal Component Analysis coupled with DNA microarray analysis indicated a shift in transcriptome from high density culture levels to those of low density culture levels following ERK inhibition. From this, it was seen that ERK inhibition resulted in the inhibition of cell cycle transcription factors.

Fig. 9

To conclude, the authors were able to demonstrate that an increase in collagen density (increase matrix stiffness) alone encouraged cellular proliferation and the development of invasive phenotypes. The mechanism by which this occurs is through the focal adhesions. Higher matrix stiffness promotes the formation of high-area focal adhesions at the epithelial-ECM interface, which in turn activates the FAK signaling pathway. FAK activates downstream signals such as Rho/ROCK and ERK to effect the formation of malignant behaviors in the MECs. An in depth biochemical analysis of the FAK cascade revealed that ERK directly interacts with more than 160 known transcription factors implicated in cell proliferation, migration, and apoptosis. Hence, ERK functions as the ‘bottleneck’ regulator of transcriptional response to physical stimuli. These results have many implications in drug development for treating breast cancer.

Significance and Feedback

Provenzano et al. sought out to answer an interesting and pressing question about breast cancer in a very logical manner. Their examination of cellular phenotype in breast tissue by means of mechanoregulation revealed a biochemical mechanism that could give rise to huge potential in the development of a cure to breast cancer. By isolating an independent risk factor (tissue density), the authors were able to carry out in depth experiments without worry of having ambiguous implications. As a result, they were able to show that an increase in matrix density led to the formation of focal adhesions. At these sites, up-regulation of FAK activates the relevant Rho/ROCK and ERK pathways, influencing the transcription of genes associated with malignancy. The elucidation of this mechanism evokes clinical relevance by illuminating several candidates for “repressing the increased risk of breast carcinoma due to high breast tissue density” (Provenzano et al., 2009).

One of the only drawbacks in this paper is the authors’ explanation of controls. The authors do not provide information on the controls done for any experiment except for the more sophisticated protocols. For instance, their explanation for using polyacrylamide gels with collagen versus just a collagen matrix to reduce ECM ligand influence proved particularly well thought out. Their data figures, however, presented reasonable negative and positive controls. One other aspect that may have been misleading was the different interpretations of low and high density matrices for transformed and non-transformed cells used in their cultures. For example, in transformed cells, densities were LD=1.3 mg/ml and HD=3.0 mg/ml while in transformed cells, densities were LD=3.0 mg/ml and HD=4.0 mg/ml (LD = low density; HD = high density).