Generation of Stable, Long-Term Cultures of Experimentally Induced Myofibroblasts from Co-Cultures of Human Mammary Fibroblasts with Breast Cancer Cells

Date of Award


Degree Type


First Advisor

Abigail Hielscher, PhD

Second Advisor

Valerie E Cadet, PhD

Third Advisor

Xinyu (Eric) Wang, PhD

Fourth Advisor

Brian Matayoshi, PhD


Background and Purpose: The breast cancer microenvironment is a significant contributor to tumor progression. Carcinoma-associated fibroblasts (CAFs) are the most predominant cell type in the microenvironment. These fibroblast-like cells promote metastasis, angiogenesis, and therapy resistance. The generation of therapies against CAFs has been delayed due to the lack of suitable CAF cultures. The purpose of this study was to generate stable long-term cultures of CAFs from multiple rounds of co-culture with human mammary fibroblasts (HMFs) and MDA-MB-231 breast cancer cells (BCC). Implications from these experiments provide further insight into mechanisms controlling the CAF phenotype, and will support the generation of a model culture to test drug therapies in the future. Experimental Approach: HMFs were co-cultured with MDA-MB-231 breast cancer cells (BCCs) in co-culture medium for 7-10 days. The transformed fibroblast population, termed experimentally induced myofibroblasts (E-Myo1), was isolated via Fluorescence Associated Cell Sorting (FACs). Immunofluorescence staining for epithelial marker Cytokeratin-19 confirmed the EMyo populations were free of BCCs. E-Myo1 cells were co-cultured with BCCs to generate E-Myo2 cells, which were co-cultured again with BCCs to generate EMyo3 cells. Extra domain A fibronectin (ED-A Fn), alpha-smooth muscle actin (α-SMA), collagen I, fibronectin (Fn), and vimentin were quantified in each population with western blot and immunofluorescence. Enzyme-linked immunosorbent assay and western blot experiments quantified the presence of signaling molecules TGF-β1, Smad2, Smad3, and phosphorylated Smad (p-Smad), that last of which have been reported to be in abundance with the myofibroblast phenotype.

Results: E-Myo cells exhibited features indicative of an activated phenotype. Purity of

each EMyo population was confirmed for EMyo1, EMyo2, and EMyo3 populations


experiments revealed that ED-A Fn was significantly (P

and EMyo3p10 populations. WB results exhibited a significant (P

Fn in EMyo2, and IF results exhibited a significant (P

EMyo3p5. In addition, WB and IF experiments revealed an increase in ED-A Fn and

fibronectin in EMyo1, EMyo2, and EMyo3p1 populations, but the increase was not

significant. WB data indicated vimentin expression increased in E-Myo cells, but α-SMA

was not different from HMF controls. Extracellular matrix (ECM) deposition of Collagen

I, ED-A Fn, and fibronectin was also increased in all EMyo populations compared to the

HMF control. TGF-β1 was significantly increased in the EMyo3p10 population


experiments of p-Smad and Smad 2/3 were performed to elucidate whether the increased

TGF-β1 increase was correlated with an increase in TGF-β signaling. Smad2/3 was

present in the EMyo and HMF populations, but the p-Smad protein was not detected in

any of the EMyo populations or control HMFs.

Conclusions: The data from this study suggests that successive co-culturing of BCCs

with HMFs results in the expression of myofibroblast markers. The isolation technique

following co-culture reported in this study presents the foundation for establishing a

protocol for the in-vitro generation of CAFs.

This document is currently not available here.