Full Professor Washington University in St. Louis, United States
Introduction: Breast cancer, characterized by uncontrollable cell growth and proliferation, is a major cause of death worldwide. According to the Breast Cancer Research Foundation, 2.3 million women were diagnosed with breast cancer and 670,000 women died in 2022 alone . The survival rate for metastasized, advanced-staged cancers is 30% 1. Hypoxia, a reduced oxygen environment, is commonly found in solid tumors. It was shown to increase extracellular matrix (ECM) remodeling and cellular motility, allowing primary tumors to metastasize3. While the interactions of hypoxia-promoted ECM remodeling and cellular motility in breast cancer are an area of high research interest, the relationship is not yet well characterized. The goals of this study are to characterize hypoxia-remodeled type I collagen and understand how these properties influence cell motility. Collagen composes more than 80% of mammary gland ECM 2 and is shown to undergo significant changes affecting the tumor microenvironment in tumor development 3. Collagen was remodeled by both tumorigenic and non-tumorigenic cells under hypoxia (~1% O2) and normoxia (standard, atmospheric ~21% O2). Single cells were seeded and tracked for 24 hr. We hypothesized that hypoxia-remodeled collagen will lead to increased cell motility and area for collagen remodeled by tumorigenic cells but remain constant for non-tumorigenic cells. Understanding the effects of hypoxia and its effects on the tumor microenvironment could possibly lead to novel methods for cancer diagnosis and treatment, leading to improved patient outcomes.
Materials and
Methods: Custom 24-well plates were created using a glass-bottomed cell culture plate (Cellvis®). A 9:1 wt. % ratio PDMS was used to fill gaps between wells. Wells were coated with 2% glutaraldehyde solution to increase collagen adhesion. The well plate was subsequently sterilized in a Krackeler Scientific UV Crosslinker . A neutralized rat-tail-derived 2mg/mL type I collagen suspension was added onto each well and allowed to crosslink. MCF10A-GFP and BT549 cells were seeded as the first round at 2×105 cells per well, two wells per group as shown in Fig. 1. Two wells were not seeded as a sham or control condition. Cells were incubated in a standard incubator under normoxia conditions or were incubated in a two-gas incubator (Eppendorf®, Galaxy 48r) maintained at a hypoxia environment (1% oxygen, determined by literature 4). After two days of incubation to allow for remodeling of collagen, cells were gently removed via trypsinization and rinsing. The second round comprised of 5000 MCF10A-GFP cells per well were seeded onto the remodeled collagen. Cells were tracked with brightfield and fluorescence microscopy for 24 hours using a Zeiss AxioObserver. Cell motility and morphology parameters were extracted using custom scripts in MATLAB, cellpose, and ImageJ. Collagen was imaged using reflectance imaging . Data analysis was performed using MATLAB. One-way T-tests and one-way ANOVA was performed using IBM SPSS (α = .05).
Results, Conclusions, and Discussions: For non-invasive breast epithelial cells (MCF10A), collagen remodeling under hypoxia was shown to increase cellular migration distance and persistence compared to normoxia (p = <.001) and no remodeling (p = <.001). Increased distance and persistence are indicative of increased collagen alignment resulting from matrix remodeling. However, greater collagen remodeling is associated with higher cell velocity; hypoxia remodeling was shown to have no effect on velocities (p = .691). Normoxia remodeling was shown to increase velocity compared to hypoxia and no remodeling (p = <.001). As shown in Fig. 2, normoxia collagen remodeling was shown to increase collagen density and decrease fiber length, allowing for increased migration velocity consistent with reported literature 5. Hypoxia collagen remodeling was shown to decrease collagen density but increase fiber length and alignment. Quantification of protein expression such as LOX or MMPs can help further explain the mechanisms affecting remodeling. For invasive, tumorigenic breast epithelial cells (BT549), hypoxia significantly changed remodeling behavior. Collagen remodeling under hypoxia was shown to increase distance traveled, velocity, and persistence compared to normoxia and no remodeling (p = <.001). Under normoxia, collagen underwent significant degradation as shown in Fig. 3B, leading to decreased velocity, distance traveled, and persistence. Under hypoxia, collagen crosslinking increased and fiber length decreased, leading to increased cell velocity and distance traveled (p = <.001). It is possible that invasive cells rely on hypoxia-induced pathways to initiate metastasis. Quantification of hypoxia-induced MMP and LOX expression and further experiments could verify this claim. Hypoxia was found to increase migration distance and persistence but only produced significant changes in collagen remodeling for tumorigenic cell lines compared to normoxia. The experiment is ultimately limited in its two-dimensional nature and should be reproduced using a three-dimensional environment. Two-dimensional studies have been shown to not consistently reproduce complex physiological phenomena in complex organisms. Further understanding how hypoxia affects the tumor microenvironment and cell phenotype would grant fundamental insights into cancer progression used to develop novel therapeutic targets for metastasis inhibition.
Acknowledgements (Optional): I would like to thank everyone from the Pathak Lab for their kindness and their constant willingness to help. I would also like to thank Hongsheng Yu, Dr. Sarah Waye, and Dr. Amit Pathak for their fantastic mentorship. I would finally like to thank The Center for Engineering and Mechanobiology for this opportunity.
