Engineered culture models for studies of tumor-microenvironment interactions

Abstract

Heterogeneous microenvironmental conditions play critical roles in cancer pathogenesis and therapy resistance and arise from changes in tissue dimensionality, cell-extracellular matrix (ECM) interactions, soluble factor signaling, oxygen as well as metabolic gradients, and exogeneous biomechanical cues. Traditional cell culture approaches are restricted in their ability to mimic this complexity with physiological relevance, offering only partial explanation as to why novel therapeutic compounds are frequently efficacious in vitro but disappoint in preclinical and clinical studies. In an effort to overcome these limitations, physical sciences-based strategies have been employed to model specific aspects of the cancer microenvironment. Although these strategies offer promise to reveal the contributions of microenvironmental parameters on tumor initiation, progression, and therapy resistance, they, too, frequently suffer from limitations. This review highlights physicochemical and biological key features of the tumor microenvironment, critically discusses advantages and limitations of current engineering strategies, and provides a perspective on future opportunities for engineered tumor models.

Figure 1

Biological and physicochemical characteristics of the tumor microenvironment. (a) Tumor initiation perturbs 2D epithelial architecture and basement membrane organization. Tumor invasion and migration toward adjacent blood vessels are promoted by increased oxygen and nutrient demands of the growing 3D tumor as well as by epithelial-to-mesenchymal transition (EMT)- and stroma-mediated changes in extracellular matrix (ECM) composition, mechanical properties, and conformation. (b) Intravasation into blood vessels introduces tumor cells into the circulation where they lose substrate adhesion and are exposed to fluid flow–mediated shear stress. (c) Lodging of tumor cell(s) in capillary beds of secondary organs (e.g., bone) facilitates their extravasation via endothelial or basement membrane protein interactions and the formation of micrometastases. Pending survival and favorable microenvironmental conditions including appropriate ECM characteristics and mechanical stimuli, secondary tumor growth ensues.

Figure 2

Engineering approaches to mimicking tumor physicochemical and biological characteristics. (a) Tumor three-dimensionality and control over cell-cell interactions can be achieved through dielectrophoretic cell patterning; cells are exposed to an electric field and migrate toward specific locations of a micropatterned electrical insulator. Adjustment of seeding density controls 3D microorganization, and gelation of the migration medium stabilizes this configuration for subsequent culture (58). (b) Culture of lung cancer cells within PEG hydrogels covalently modified with RGD adhesion peptides and MMP-sensitive cross-linkers undergo morphogenetic and polarity changes similar to those that occur in Matrigel^® (98). (c) Integrating microfluidic channels into hydrogels affords temporal and spatial control of soluble factor signaling by adjusting perfusion rate and modular network assembly, respectively (122). (d) Lung-on-a-chip microdevices mimic respiratory mechanics for possible future studies of their effects on lung-metastatic tumor cells (166). These devices recreate physiological breathing movements by applying vacuum to the side chambers, which causes mechanical stretching of a porous PDMS membrane seeded with lung epithelial cells on the top and endothelial cells on the bottom. Membrane stretching creates cellular tension in the direction of the applied force. Abbreviations: DCP, dicalcium phosphate; ECM, extracellular matrix; ITO, indium tin oxide; MMP, matrix metalloproteinase; PDMS, polydimethylsiloxane; PEG, poly(ethylene glycol); RGD, Arg-Gly-Asp. (Images were modified and included with permission from Nature Publishing Group, American Association of Cancer Research, and American Association for the Advancement of Science.)

Figure 3

Multiscale nature of cancer contributes to its complexity. Cancer represents a multiscale disease in which the integrated effects of molecular-, cellular-, tissue-, and organ-level signaling cause systemic disease that is characterized by tremendous complexity. Yet, most technologies focus on recapitulating signaling events at a single scale. Specific examples include (from left to right) analysis of fibronectin molecular conformation via FRET imaging to evaluate how tumors modulate ECM unfolding (21), cell patterning techniques to investigate signaling responses of individual cells and clusters of cells (51), 3D PLGA scaffold culture to recapitulate histological and functional characteristics of tumors in vivo (4), and combination of decellularized organs with bioreactor technologies such as artificial ventilation models to study cancer cell behavior within a fully functioning organ (e.g., the lung) (173). Integrating such single-scale models into more complex cell culture analogs that allow interconnecting various compartments similar to previously developed body-on-a-chip models (172) would be useful to mimic tumor complexity. Such systems would not only enable studies of cell metastasis to secondary locations or evaluate the toxicity of novel anticancer therapeutics but also test the effect of other conditions such as obesity or inflammation on tumorigenesis. Abbreviations: 3D, three dimensional; ECM, extracellular matrix; FRET, fluorescence resonance energy transfer; PLGA, poly(lactic-co-glycolic acid). (Photographs were included with permission from Nature Publishing Group, Royal Society of Chemistry, Annual Reviews, and Institute of Physics Publishing.)