Extracellular matrix cues combine to generate diverse phenotypes during breast cancer progression

Introduction:: Breast cancer is the second leading cause of death among women globally and approximately 13% of women develop invasive breast cancers during their lifetime. Invasive ductal carcinomas are associated with changes in composition and mechanical properties of the extracellular matrix, such as increased stiffness, altered stress relaxation rate, and increased production of collagen-I. While reductionist models have been useful in understanding the role of individual microenvironmental cues in promoting malignant traits, these properties vary simultaneously in vivo. How the cell weighs each individual factor during phenotypic determination remains unknown, in part due to the lack of platforms in which these three parameters can be independently tuned. To answer this, we have developed an alginate-based 3D hydrogel cell culture platform with independently tunable stiffness, stress relaxation, and collagen-I concentration to map a range of cellular microenvironments.


Materials and Methods:: We encapsulated three types of mammary epithelial cell lines showing varying degrees of malignancy specifically, MCF10A (non-tumorigenic), MCF7 (tumorigenic but non-metastatic), and MDA-MB-231 (metastatic) into alginate-collagen and cultured them for 7 days. Calcium ion concentration in these matrices was tuned to produce soft (Young’s modulus ≈ 100 Pa ) and stiff (Young’s modulus ≈ 2000 Pa ) matrices. Molecular mass of the alginate was varied to produce slow relaxing (t1/2 ≈ 1000 s ) and fast relaxing (t1/2 ≈ 100 s ) matrices. Finally, the collagen-I incorporated in these matrices varied from low (0.5 mg/ml) to high concentration (2 mg/ml) to tune the fiber density. Confocal microscopy was used to image single cells and spheroids, and shape metrics were used to quantify morphology. . EdU positive cells were counted to assess cell proliferation. RNA-seq was performed on MDA-MB-231 cells to determine differentially expressed genes.


Results, Conclusions, and Discussions:: We were able to independently modulate the stiffness, stress relaxation, and collagen density of our alginate-collagen I hydrogels. Following 7 days of culture, we analyzed changes in cell shape and observed that cells encapsulated in soft matrices form significantly rounder clusters than in stiff matrices, indicating a less-invasive phenotype. Also, cell clusters in fast relaxing matrices were more invasive than in slow relaxing matrices. Strikingly, the invasiveness and morphological changes induced by fast relaxing matrices occur to similar extents in both soft and stiff matrices and across all collagen concentrations, and are thus independent of other matrix parameters. Increasing the collagen density also led to an increase in invasiveness for cells in slow relaxing matrices, and this was further enhanced by high stiffness. To validate these findings in a more physiologically relevant invasion assay, we encapsulated spheroids in these matrices and found a comparable trend in spheroid invasion as the single cell studies. Consistent with cells displaying enhanced malignant traits, we observed that high stiffness, faster stress relaxation, and high collagen density promoted cell proliferation in these matrices. KEGG pathway analysis on our preliminary RNA-seq data demonstrated cue-specific pathway enrichment in response to individual ECM cues. Currently, we are investigating how combination of various ECM cues leads to the enrichment of novel gene sets. In summary, our 3D hydrogel system with independently tunable stiffness, stress relaxation, and collagen density produces an array of phenotypes in breast cancer cells. We observe that under certain combinations of cues, one cue may dominate over the others in dictating cell behavior. In the future, such a platform can also help us identify potentially novel targets that enable cancer cells to sense and respond to ECM cues.


Acknowledgements (Optional): :

References (Optional): :