A cell-instructive hydrogel to regulate malignancy of 3D tumor spheroids with matrix rigidity

Abstract

Three dimensional (3D) tumor spheroid models are becoming important biomedical tools for both fundamental and applied cancer studies, but current models do not account for different levels of cancer malignancy. Several studies have reported that the mechanical rigidity of a hydrogel plays a significant role in regulating the phenotypes of cancer cells adhered to the gel surface. This finding suggests that matrix rigidity should also modulate the malignancy of 3D tumor spheroids. However, the role of matrix stiffness is often confounded by concurrent changes in 3D matrix permeability. This study reports an advanced strategy to assemble 3D liver tumor spheroids with controlled intercellular organization, phenotypes, and angiogenic activities using hydrogels with controlled stiffness and minimal differences in molecular diffusivity. The elastic moduli of cell-encapsulated collagen gels were increased by stiffening interconnected collagen fibers with varied amounts of poly(ethylene glycol) di-(succinic acid N-hydroxysuccinimidyl ester). Interestingly, hepatocellular carcinoma cells encapsulated in a fat-like, softer hydrogel formed malignant cancer spheroids, while cells cultured in a liver-like, stiffer gel formed compact hepatoids with suppressed malignancy. Overall, both the hydrogel and the 3D tumor spheroids developed in this study will be greatly useful to better understand and regulate the emergent behaviors of various cancer cells.

Introduction

Cancer progression is largely determined by genetic mutations and microenvironmental changes [1]. Great advances in genomic analysis have contributed to identification and understanding of key genetic mutations in cancer progression. In contrast, there has been only modest progress in understanding the role of the tumor microenvironment in aberrant gene expression and the resulting phenotypic changes of cancer cells, because of our limited ability to replicate the cancer microenvironment.

Therefore, efforts have been increasingly made to assemble in vitro 3D tumor models with structural similarity to in vivo cancers of interest [2], [3], [4], [5]. One popular 3D tumor model is comprised of tumor spheroids, which are commonly prepared by culturing cells suspended in spinner flasks or plated on non-adhesive agar-coated plates [4], [5]. However, the resulting spheroids are formed by forced aggregation of cancer cells and largely present malignant phenotypes. This makes it difficult to examine the potential effects of the tumor microenvironment on cancer cells’ phenotypic changes.

Recently, several studies reported that the mechanical rigidity of a collagen gel, largely controlled by varying the collagen concentration, plays a pivotal role in regulating the malignancy of cancer cells adhered to gel surfaces [6]. This finding suggests that the matrix stiffness would also tune the intercellular organization and malignancy level of cancer cells cultured in a 3D collagen gel. However, this approach to control gel stiffness with the collagen concentration is accompanied by changes in gel permeability and number of cell adhesion sites, both of which potentially influence cellular activities and viability and obscure the role of matrix rigidity [7], [8].For example, increasing gel stiffness solely by increasing the collagen concentration also results in a decrease in gel permeability but an increase in the number of cell adhesion sites. Earlier studies have demonstrated that these particular changes cause a decrease in the growth rate and viability of cells within the 3D gel matrix. In addition, certain approaches to control stiffness with a chemical or enzymatic cross-linker are plagued by the cytotoxicity of the chemical cross-linker and a limited controllable range of stiffness [7], [9]

This study therefore presents a strategy to control the stiffness of a cell-encapsulating fibril collagen hydrogel while minimizing changes in gel permeability and the number of cell adhesion sites. We further used the resulting hydrogel to build 3D advanced liver tumor spheroids with controlled intercellular organization, cell proliferation, metabolite-detoxification, and tumor vessel growth. In this study, the gel stiffness was controlled by forming supplemental chemical cross-links between self-associated collagen fibrils with varied concentrations of biocompatible poly(ethylene glycol) di-(succinic acid N-hydroxysuccinimidyl ester) (PEG-diNHS) (Scheme 1). The resulting collagen-PEG gel stiffness was evaluated by measuring the compressive elastic modulus. In parallel, the gel architecture was examined by imaging the collagen fibrils with the second-harmonic generation (SHG)-based confocal imaging technique. Finally, the gel permeability was determined by measuring the diffusion coefficient of a fluorescent macromolecule using fluorescence recovery after photobleaching (FRAP).

The role of gel stiffness in controlling the structure and phenotype of tumor spheroids was evaluated by encapsulating hepatocarcinoma cells (HCCs) in hydrogels with varied stiffness. The effect of hydrogel stiffness on the assembly of 3D tumor spheroids was evaluated by examining the intercellular organization, phenotype, and angiogenic activities of resultant spheroids. Changes in the intercellular organization of resultant spheroids were assessed by characterizing the sizes of the cell nucleus and spheroid and the expression levels of β1-integrin and E-cadherin. The phenotypic changes in the spheroids were evaluated by analyzing cell growth rates and hepatocyte-specific metabolic activities. Finally, the spheroids’ proangiogenic activities were examined by measuring the cellular production of vascular endothelial growth factor (VEGF) and evaluating blood vessel formation on a chicken chorioallantoic membrane (CAM).