1. Introduction
Three-dimensional cell culture models such as multicellular spheroids and organoids have been demonstrated to mimic several biological processes in vitro much better than monolayer cell cultures (Cui, Hartanto, & Zhang, 2017). There are essentially two methods to produce uniform-size multicellular spheroids: scaffold free, in which the cells are prevented from adhering to substrates but not to each other thus forming spheroids, and scaffold- or matrix-based, in which cells are embedded into a three-dimensional (3D) biomaterial such as a hydrogel. The latter, usually composed of polymers, has the advantage to provide a physical structure comparable, in terms of stiffness and viscoelasticity, to the extracellular matrix (ECM) in vivo and it is thus able to reproduce not only cell-cell interactions, but also cell-matrix interplays directly related to phenomena associated with cell growth (cell fate) such as cytoskeletal organization, gene expression, nutrients diffusion, pH and hypoxia (Singh, Brito, & Lammerding, 2018). This makes matrix-based spheroids increasingly important for studying solid tumors microenvironment and the response to drug treatments (Nath & Devi, 2016).
Among several 3D biomaterials, hydrogels of complex biological origin have been used in many studies. The most common are MatrigelTM, derived from Engelbreth-Holm-Swarm mouse tumor sarcoma, and other basement membrane-rich matrices. These hydrogels contain matrix membrane proteins, hormones and soluble growth factors whose composition may vary among different batches. This aspect limits their use because, although they allow for the growth in three dimensions, the batch-to-batch variability could potentially alter cell culture systems and experimental results. Thus, other simple hydrogels, without hormones and growth factors, able to maximize reproducibility and offering the possibility to tune biochemical as well as mechanical properties, have been considered to grow spheroids. Hydrogels with these features are made of natural, synthetic or hybrid materials, such as alginate, agarose, collagen, hyaluronic acid and polyethylene glycol (Cui et al., 2017; Teruki, Kimiko, & Yasuhiko, 2020).
Among them, agarose is an inert, inexpensive, and easily available polysaccharide derived from red marine algae and consisting of repeated units of b-1,3-linked-D-galactose and a-1,4-linked 3,6-anhydro-L-galactose. It possesses excellent biocompatibility, optimal gelling features and tunable mechanical properties that boosted its use as biomaterial for the manufacturing of tissue engineering scaffolds (Anderson & Johnstone, 2017; López-Marcial et al., 2018). In addition, thanks to the ease of preparation, it has been recently exploited as non-adhesive and micromolded substrate for the growth of tumor spheroids based on multicellular aggregation (Napolitano et al., 2007; Tang, Liu, & Chen, 2016). Nevertheless, due to its poor bioadhesivity, as it lacks cell adhesion motifs, its use as ECM-mimicking material is very limited.
On the other hand, type I collagen is the main protein component of the ECM in mammals. The presence of cellular binding sites (i.e. the “GxOGER” sequence, where “R” is arginine, “G” is glycine, “D” in aspartate, “O” is hydroxyproline, “E” is glutamate and “x” is a hydrophobic amino acid) that promote cell adhesion and proliferation, and regulate cell signaling pathways, makes collagen highly bioactive and suitable for the development of natural material-based hydrogels for cell culture studies (Davidenko et al., 2018; Davidenko et al., 2016; Hamaia & Farndale, 2014; Tibbitt & Anseth, 2009). In addition, it plays a crucial role in tumor progression and invasion (Li & Kumacheva, 2018). It can form hydrogels through self-aggregation of the fibers or by physical and chemical crosslinking, but is suffers of poor mechanical properties, limited stability over time and high costs (Caliari & Burdick, 2016).
In this sense, blended hydrogels composed of agarose and collagen combine the mechanical properties of agarose and the biomimetic nature of collagen (Ulrich, Jain, Tanner, MacKay, & Kumar, 2010). As far as we know, only one study described the use of hydrogels containing agarose and collagen as matrices for the formation of tumor spheroids. In detail, composite hydrogels made of alginate, marine collagen, and agarose were prepared and the growth and viability of spheroids derived from ovarian cancer and lymphoma cell lines were monitored up to 14 days (Shin et al., 2016). Notably, twoother studies described the use of agarose-collagen blend hydrogels that were not used to promote the growth of spheroids, but to unravel the effects of biophysical cues on cellular mechanobiology of 2D intervertebral disc cells(Cambria et al., 2020) and glioblastoma spheroids that had been previously prepared by the hanging drop method and then loaded into the hydrogel (Ulrich et al., 2010). Both works evidence the crucial role of the hydrogel stiffness and adhesivity as driving forces that modulate the cell plasticity and connect the biological functionality to the surrounding physical stimuli. Thus, the interplay between the hydrogel and the cells embedded within it depends on what the cells sense and how they reply to the host environment. In living tissues, the stiffness spans from few tens Pa in intestinal mucus to GPa of bones, and variations of the mechanical properties are typically associated to diseases and cancer development (Guimarães, Gasperini, Marques, & Reis, 2020). In the case of breast, it has been shown that the stiffness increases from hundreds Pa to few kPa when the normal tissue undergoes to tumor transformation, due to the remodeling of the ECM (Butcher, Alliston, & Weaver, 2009).
Therefore, artificial matrices resembling the physical properties of the naïve tumor environment may facilitate the study of 3D tumors behavior in vitro and predict their response to modifications of the mechanical cues or to drug treatments, as it occurs during cancer progression and metastatization (Singh et al., 2018).
In this work we have developed agarose-collagen blended hydrogels with variable agarose amount (from 0.5 to 0.125%) and analyzed the growth of breast tumor spheroids of three different breast cancer cell lines: the luminal estrogen receptor positive cells, MCF-7 and MDA-MB-361, and the triple negative model, MDA-MB-231, for comparative analysis.
The preparation method is simple, quick and highly reproducible. By embedding individual cells into the liquid mixture of agarose and collagen, prior to the gelation, allows to follow up the growth of the 3D tumoroids over time. The physical properties of the matrices have been thoroughly investigated and the characteristics of the spheroids’ growth were related to the hydrogel features. Upon variation of the agarose percentage, the physical and mechanical properties of the matrices can be tuned, and in turn, the growth kinetic and the morphology of the spheroids also vary, confirming their sensitivity to the environment stiffness. Noteworthy, the optical transparency of agarose facilitates the daily analysis of spheroids that result more compact in stiffer hydrogels, while they look loosened with cellular protrusions and disseminated cells in softest matrices. Remarkably, tumor spheroids can be recovered upon enzymatic treatment of the hydrogels with agarase, with a yield of 100% in the case of the matrices containing the lower amount of agarose (equal to 0.125%). Preliminary drug testing studies with cisplatin evidenced a cytotoxic response that depends on the agarose percentage concentration. Overall, the results show that agarose-collagen blended hydrogels are suitable modular matrices for tumor spheroids culturing and for evaluating drug protocols efficacy.