Abstract Robust, scalable, and quantitative in vitro tumor models that recapitulate the complex multicellular in vivo tumor microenvironment are lacking. Most in vitro tumor models fail to recapitulate the abnormal tissue architecture characteristic of solid tumors, which has been shown to promote tumor progression through forced-depolarization, enhanced cell-cell contacts, and loss of tensional homeostasis. Here, we used an in vitro tumor model that effectively mimics the in vivo multicellular architecture, which we refer to as the 3D collagen embedded spheroid model, to quantitatively investigate the response of breast cancer cells to chemotherapeutic intervention and assess the presence of the highly tumorigenic subpopulation of cancer cells known as cancer stem cells (CSCs). We compared our embedded spheroid model to two other in vitro models—a 2D monolayer culture and a 3D diffusely embedded single-cell model—and show that our 3D embedded spheroid model exhibits a more robust drug response as well as an enriched CSC subpopulation. In this study, we investigated the efficacy of two front-line chemotherapeutics (paclitaxel and cisplatin) against cells from the post-metastatic and triple-negative breast cancer cell line MDA.MB.231 cultured within the aforementioned models. In the 3D diffuse model, cells were embedded within 4 mg/mL collagen gels at a seeding density of 105 cells/mL. For the embedded spheroid model, spheroids composed of 104 cells were embedded within a 4 mg/mL collagen gel. The treatment regimen for all models consisted of 72 hours of drug exposure—10 ng/mL for paclitaxel and 1.5 μg/mL for cisplatin—followed by removal of drug and an additional 72 hours of culture. Upon completion of the experiment, cells were harvested and drug efficacy was quantified via cell viability measurements using an MTS assay. Moreover, for the embedded spheroid condition, collagenase treatment was used to isolate two distinct populations: 1) cells that remained within the spheroid structure, termed the “core population”; and 2) cells that had invaded into the surrounding collagen, termed the “periphery population”. Our results showed that both drugs effectively reduced viability to about 15% in the 2D monolayer, and 30% in the 3D diffusely embedded and 3D spheroid periphery conditions. However, cells in the core population exhibited a highly robust response against the agents with significantly higher viabilities of 91 and 97% for paclitaxel and cisplatin, respectively. Additionally, we sought to quantify the CSC content across the three models to determine whether the aforementioned differences in drug efficacy correlated with an enriched CSC content. Here, we used three independent techniques to quantify CSC content within our untreated conditions: 1) Aldefluor assay, 2) Mammosphere assay, and 3) TaqMan RT-qPCR assay of two known CSC markers (ALDH1A3 and SOX2). The Aldefluor assay revealed CSCs to be preferentially located in the spheroid core compared to the spheroid periphery, which was independently corroborated by both the mammosphere and RT-qPCR assays. In addition, the high sensitivity of the TaqMan assay provided us with the means to assess ALDH1A3 and SOX2 expression across our in vitro models following treatment with either paclitaxel or cisplatin. Specifically, treatment with either paclitaxel or cisplatin led to a statistically significant increase in ALDH1A3 and SOX2 expression across all models, compared to the untreated 2D monolayer. These findings indicate that paclitaxel and cisplatin efficacy is inversely related to CSC content and that these therapies can even enrich for the highly malignant CSC subpopulation. In conclusion, the 3D embedded spheroid model described herein recreates a phenotypic landscape possessing an enriched and spatially-dependent CSC population that reflects the response of CSCs to clinical treatment. Citation Format: Daniel S. Reynolds, Kristie M. Tevis, Mark W. Grinstaff, Muhammad H. Zaman. Chemotherapeutic treatment enriches for cancer stem cell content in breast cancer spheroids. [abstract]. In: Proceedings of the AACR Special Conference on Engineering and Physical Sciences in Oncology; 2016 Jun 25-28; Boston, MA. Philadelphia (PA): AACR; Cancer Res 2017;77(2 Suppl):Abstract nr A19.
Read full abstract