Chemoresistance of Cancer Cells: Requirements of Tumor Microenvironment-mimicking In Vitro Models in Anti-Cancer Drug Development

Abstract

For decades, scientists have been using two-dimensional cell culture platforms for high-throughput drug screening of anticancer drugs. Growing evidence indicates that the results of anti-cancer drug screening vary with the cell culture microenvironment, and this variation has been proposed as a reason for the high failure rate of clinical trials. Since the culture condition-dependent drug sensitivity of anti-cancer drugs may negatively impact the identification of clinically effective drug candidates, more reliable in vitro cancer platforms are urgently needed. In this review article, we provide an overview of how cell culture conditions can alter drug efficacy and highlight the importance of developing more reliable cancer drug testing platforms for use in the drug discovery process. The environmental factors that can alter drug delivery and efficacy are reviewed. Based on these observations of chemoresistant tumor physiology, we summarize the recent advances in the fabrication of in vitro cancer models and the model-dependent cytotoxicity of anti-cancer drugs, with a particular focus on engineered environmental factors in these platforms. It is believed that more physiologically relevant cancer models can revolutionize the drug discovery process.

Keywords: biomimetic; cancer cell; chemoresistance; efficacy; tumor microenvironment.

Figure 1

The differences between the native tumor microenvironment (TME) and the conventional in vitro cancer models in terms of the recapitulation of physiological factors. (A) The physiological conditions within the native TME. (B) The features of the conventional 2D or plastic dish-based cancer models. Because the conventional cancer models do not reflect the important environmental cues observed in the TME, the behaviors and responses of cancer cells in vivo cannot be fully recapitulated in the experimental conditions. In particular, tests of the efficacy or cytotoxicity of anticancer drugs frequently show misleading drug screening results, increasing the time and cost of drug discovery.

Figure 2

The tumor microenvironmental factors that cause chemoresistance of cancer cells. Physical cues include the physical barrier, binding to the extracellular matrix component, stiffness-induced mechanotransduction, and fluidic shear stress. Biological and biochemical cues include hypoxia, low pH, cell-cell interaction, cancer-associated fibroblasts, and tumor-associated macrophages. Because each cue induces the chemoresistance of cancer cells through different mechanisms, a combinatorial consideration of those factors using innovative in vitro cancer models is required to identify the exact efficacy of anticancer drugs.

Figure 3

The tumor spheroid as a three-dimensional model for recapitulating the environmental cues originating from the high density of cancer cells such as a physical barrier against drug delivery, the concentration gradient of oxygen and nutrition, and low pH in the core of the tumor mass. (A) Representative features of tumor spheroids in terms of the physiological similarity with the native tumor. (B-a) The response of prostate cancer cell (LNCaP) spheroids in the aggregated form against the anticancer drug docetaxel. (B-b) The difference in cell viability depending on the culture dimension. Reprinted with permission from Chambers et al.. (C-a) The time-dependent penetration of doxorubicin into the tumor spheroids. Red: doxorubicin. (C-b) The size-dependent chemoresistance of human breast adenocarcinoma spheroids against doxorubicin. Reprinted with permission from Gong et al..

Figure 4

Engineered platforms that mimic the fluidic shear stress around native tumor cells. (A) Schematic illustration of tumor spheroids and adhered cancer cells in a 3D matrix or circulating tumor cells upon exposure to a physiologically relevant stream. (B-a) A photograph and schematic illustration of the tumor spheroid-embedding microfluidic device. (B-b) The viable cells (denoted as Annexin V^-/PI^- cells) after the treatment with anticancer drugs in the presence or absence of fluidic shear stress. Reprinted with permission from Ip et al.. (C-a) The microscopic images of COLO 205 cells exposed to tumor necrosis factor apoptosis-inducing ligand (TRAIL) and fluidic shear stress. Scale bar = 30 μm. (C-b) Apoptosis of COLO 205 cells by TRAIL in the presence or absence of fluidic shear stress. The COLO 205 cells were treated with 0.1 μg/mL TRAIL before applying fluidic shear stress. *p<0.05. Reprinted with permission from Mitchell et al.. (D-a) Fluorescence images of the sectioned scaffolds with Ewing sarcoma cells. Blue: DAPI. Scale bar = 200 μm. (D-b) The released amount of IGF-1 in response to fluidic shear stress (S: static, B-04: 0.04 mL/min, B-08: 0.08 mL/min, and B-40: 0.4 mL/min). After the treatment with the IGF-1R inhibitor (MK-0646), the release of IGF-1 was reduced in all fluidic conditions, whereas the addition of exogenous IGF-1 significantly promoted the amount of released IGF-1. Reprinted with permission from Santoro et al..

Figure 5

Multiscale topography- and 3D scaffold-induced cancer cell behaviors that mimic the physical aspect of the cell-stroma interactions. (A) Schematic illustration of cancer cells (either spheroids or single cell) in response to multiscale topography and 3D scaffolds. (B-a) Preparation of the cell morphology-imprinted substrates. Two types of cell topography-replicated substrates (negative and positive) are prepared as well as a flat one for control. (B-b) The surface topography-dependent viability of the Ishikawa cells treated with paclitaxel and doxorubicin for 48 h. Cancer cells showed different sensitivities depending on the topography, showing the highest sensitivity in the negative topography. Reprinted with permission from Tan et al.. (C-a) Scanning electron microscope images of Ewing sarcoma cells cultured on a 3D electrospun poly(ε-caprolactone) fiber matrix. Scale bar: upper row = 200 μm and lower row = 50 μm. (C-b) The response of Ewing sarcoma cells to doxorubicin. Percentage of cell viability for the 2D and 3D culture conditions, and percentage of tumor volume for in vivo as xenografts. *p<0.05 for 3D vs. 2D. Reprinted with permission from Fong et al.. (D-a) The 3D scaffolds that can incorporate dissociated tumor cells and tumor spheroids. (D-b) Efficacy of doxorubicin against U251 cancer cells depending on the culture conditions. MS: 3D scaffolds with monolayer-cultured cells. SS: 3D scaffolds with spheroid-cultured cells. Reprinted with permission from Ho et al.