Programming temporal stiffness cues within extracellular matrix hydrogels for modelling cancer niches

Abstract

Extracellular matrix (ECM) stiffening is a common occurrence during the progression of many diseases, such as breast cancer. To accurately mimic the pathophysiological context of disease within 3D in vitro models, there is high demand for smart biomaterials which replicate the dynamic and temporal mechanical cues of diseased states. This study describes a preclinical disease model, using breast cancer as an example, which replicates the dynamic plasticity of the tumour microenvironment by incorporating temporal (3-week progression) biomechanical cues within a tissue-specific hydrogel microenvironment. The composite hydrogel formulation, integrating adipose-derived decellularised ECM (AdECM) and silk fibroin, was initially crosslinked using a visible light-mediated system, and then progressively stiffened through spontaneous secondary structure interactions inherent between the polymer chains (∼10-15 kPa increase, with a final stiffness of 25 kPa). When encapsulated and cultured in vitro, MCF-7 breast cancer cells initially formed numerous, large spheroids (>1000 μm² in area), however, with progressive temporal stiffening, cells demonstrated growth arrest and underwent phenotypic changes resulting in intratumoral heterogeneity. Unlike widely-investigated static mechanical models, this stiffening hydrogel allowed for progressive phenotypic changes to be observed, and fostered the development of mature organoid-like spheroids, which mimicked both the organisation and acinar-structures of mature breast epithelium. The spheroids contained a central population of cells which expressed aggressive cellular programs, evidenced by increased fibronectin expression and reduction of E-cadherin. The phenotypic heterogeneity observed using this model is more reflective of physiological tumours, demonstrating the importance of establishing temporal cues within preclinical models in future work. Overall, the developed model demonstrated a novel strategy to uncouple ECM biomechanical properties from the cellular complexities of the disease microenvironment and offers the potential for wide applicability in other 3D in vitro disease models through addition of tissue-specific dECM materials.

Graphical abstract

Fig. 1

Schematic of tumour progression, highlighting dynamic changes in the structure and mechanics of the extracellular matrix (ECM) which modulate tumour cell behaviour. Throughout tumour progression, changes in deposition, alignment, composition and crosslinking of the ECM result in increasing tumour stiffness.

Fig. 2

Crosslinking methods to form AdECM and silk hydrogels. Visible light-mediated crosslinking with ruthenium (Ru) and sodium persuflate (SPS) forms dityrosine bonds between materials, and spontaneous hydrogen bonding forms β-sheets between silk fibroin fibrils over time without external stimuli.

Fig. 3

Overview of the breast cancer model characteristics and experimental workflow. The AdECM and silk fibroin composite hydrogel contains two natural materials which provide tissue-specific ECM cues and temporal biomechanical cues, which modulate tumour cell behaviour. The developed hydrogel assesses the importance of matrix stiffening and ECM composition in driving tumour progression without coculturing multiple cell types.

Fig. 4

Development of AdECM from porcine adipose tissue. a. Visual overview of the decellularisation protocol involving step-wise removal of cellular material and lipids using SDS and isopropanol, respectively. The resulting decellularised material was lyophilised and ground to a powder before digestion with pepsin. b. Photopolymerisation of AdECM using Ru/SPS and visible light (400–450 nm).

Fig. 5

Compositional characterisation of AdECM. a. Histological evaluation of tissue structure – H&E, collagens – Masson's Trichrome, and glycosaminoglycans (GAGs) – Alcian Blue compared with native adipose tissue. b. Biochemical evaluation of collagen, GAG and DNA contents compared with native adipose tissue. Error bars denote ± standard deviation of the mean. c. Relative abundance of core matrisome and matrisome-associated proteins detected in AdECM. d. Relative abundance of matrisome proteins, categorised by protein subgroup (colours as per panel c)). Data is presented in log scale, where low and high abundance proteins are presented on the left and right side of the baseline, respectively. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Protein structural conformation and mechanical properties of AdECM/silk hydrogel compositions. β-Sheet crystallinity of a. native, and b. methanol-treated hydrogels. c. Mean diameter and d. compressive modulus (kPa) of native and methanol-treated hydrogels. Dotted line shows size of casting mould. Error bars represent ±standard error. Asterisks denote significant differences, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001.

Fig. 7

Longitudinal changes in mechanical and structural properties of AdECM/silk hydrogel compositions. a. Compressive modulus (kPa) and b. mean diameter of hydrogels incubated in DMEM over 21 days. Error bars represent ±standard error. Asterisks denote significant differences at each time point, *p < 0.05, ***p < 0.001. c. Macro images of hydrogels at day 0, 7 and 21.

Fig. 8

MCF-7 breast cancer cell encapsulation within AdECM/silk hydrogels. a. Cell viability 24 h after encapsulation within biomaterials. b. Metabolic activity and c. compressive modulus over 21 days in culture. d. Diameter of cell-laden hydrogels after 21 days. e. Representative images of viability staining after 24 h. Calcein-AM (green) and propidium iodide (red) were used to stain live and dead cells respectively. Error bars represent ±standard error. Asterisks denote significant differences at each time point, *p < 0.05, **p < 0.01, ****p < 0.0001. See Table S1and Table S2 for statistical details. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 9

MCF-7 spheroid phenotype progression over 21 days in AdECM/silk hydrogels. a. Average spheroid size, b. number of spheroids per field-of-view, and c. representative immunofluorescence images of actin (red, phalloidin) and nuclei (blue, DAPI). White arrows indicate cell protrusions and the compressive modulus of each hydrogel composition at each time point is noted. Representative images are provided from three replicate biological experiments, each containing three hydrogel technical replicates with triplicate fields-of-view obtained from each. Error bars represent ±standard error. Asterisks denote significant differences, *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)

Fig. 10

Assessment of MCF-7 spheroid structure and ECM dynamics in AdECM/silk hydrogels. a. Representative images at 20 × magnification of haematoxylin and eosin stained sections at day 7 and day 21. The compressive modulus of each hydrogel composition is noted. Representative images are provided from three replicate biological experiments, each containing three hydrogel technical replicates with triplicate fields-of-view obtained from each. b. Quantification of matrix metalloproteinase (MMP) and tissue inhibitor metalloproteinase (TIMP) enzymes secreted at day 1 and day 21. c. Fold-change in MMP and TIMP expression between day 1 and day 21.

Fig. 11

Epithelial-mesenchymal transition and extracellular matrix production of MCF-7 cells in AdECM/silk hydrogels. a. E-cadherin, b. collagen I and c. fibronectin expression over 21 days in culture. Error bars represent ±standard error. Asterisks denote significant differences at each time point, **p < 0.01, ****p < 0.0001. See Table S3, Table S4 and Table S5 for statistical details. d. Representative immunohistochemistry images of E-cadherin, N-cadherin, collagen I and fibronectin. The compressive modulus of each hydrogel composition is noted. Representative images are provided from three replicate biological experiments, each containing three hydrogel technical replicates with triplicate fields-of-view obtained from each.