When researchers test potential cancer treatments in the laboratory, they traditionally grow cancer cells in a flat dish, a method known as 2D cell culture. While simple and inexpensive, this approach has a critical limitation: it does not reflect how cancer cells actually behave inside a living body. Cancer cells in a flat dish lack the surrounding support cells, blood vessels, and structural proteins that surround tumors in real tissue.
This matters enormously because the tumor microenvironment, the complex ecosystem around and within a tumor, plays a major role in determining how a drug works. Cancer-associated fibroblasts stiffen the tissue, macrophages can suppress or promote immune responses, and blood vessels control how much of a drug actually reaches cancer cells. When these factors are absent in a dish, drugs often appear more effective than they truly are, contributing to the high failure rate of promising compounds when they reach human clinical trials.
Colorectal cancer (CRC) is the third most common cancer worldwide, and despite advances in treatment, survival rates for advanced disease remain poor. Better preclinical models that more accurately predict drug behavior could help identify effective treatments faster and reduce the costly failures that slow progress in oncology.
Gelatin methacrylate (GelMA) is a biomaterial derived from gelatin, the same protein found in bone and skin. When mixed with a photoinitiator and exposed to UV light, GelMA cross-links to form a soft, flexible hydrogel, meaning a gel that holds a large amount of water. This water-swollen structure closely mimics the physical properties of the extracellular matrix (ECM), the natural protein scaffold that surrounds cells in body tissue.
GelMA hydrogels can be adjusted in stiffness, porosity, and degradation rate by changing their concentration. Cells embedded in or grown on GelMA form three-dimensional structures that more closely resemble real tissue than flat culture dishes. Importantly, GelMA supports the growth of multiple cell types simultaneously, making it ideal for building models that include both cancer cells and the supporting cells of the tumor microenvironment.
In this study, researchers used GelMA hydrogels to construct a multicellular 3D colorectal cancer model that included four key cell types found in real CRC tumors: Caco-2 colorectal cancer cells, cancer-associated fibroblasts (CAFs), M2 macrophages (a type of immune cell that suppresses immune responses), and HUVEC endothelial cells (which form blood vessel walls). This combination recapitulates the core structural and biological features of a real colorectal tumor.
The researchers prepared GelMA hydrogels at three different concentrations (3%, 5%, and 7%) and characterized their physical properties, including stiffness, porosity, degradation rate, and ability to support cell survival. The 5% GelMA formulation was selected as optimal, providing the right balance of structural integrity and biological compatibility for the CRC model.
To build the vascular component of the model, human umbilical vein endothelial cells (HUVECs) were grown within the GelMA hydrogel, supported by angiogenic growth factors (VEGF, Angiopoietin-1, and Angiopoietin-2). The endothelial cells formed tube-like structures within the gel, mimicking the small capillaries (blood vessels) that deliver oxygen and nutrients to a real tumor.
The full multicellular model was assembled by first seeding CAFs and M2 macrophages on the culture well surface, then adding the vascularized hydrogel on top, and finally attaching Caco-2 CRC cells to the surface of the hydrogel. This layered architecture means the cancer cells experience physical and chemical signals from fibroblasts, macrophages, and vessel-like structures, just as they would in a real tumor.
5-fluorouracil (5-FU) is one of the most widely used chemotherapy drugs for colorectal cancer. To test whether the 3D model behaved differently from standard 2D cultures, researchers treated both systems with the same concentrations of 5-FU and measured several outcomes: cell death (cytotoxicity), DNA damage (genotoxicity), oxidative stress (reactive oxygen species, or ROS), and programmed cell death (apoptosis).
The results clearly showed that cancer cells in the 3D GelMA model were significantly more resistant to 5-FU than those in 2D flat cultures. At lower drug concentrations and after longer treatment periods, the 3D cells showed notably reduced cytotoxicity, lower ROS production, and lower levels of apoptosis compared to 2D cells exposed to the same drug dose. This resistance is consistent with what is seen clinically, where some colorectal cancers do not respond well to 5-FU.
The researchers attribute the reduced drug effect in 3D to two main factors: the multicellular interactions between cancer cells, fibroblasts, macrophages, and endothelial cells (which actively signal and protect tumor cells), and diffusion-mediated effects, meaning the gel matrix slows drug penetration, so inner cancer cells receive lower drug concentrations. This is exactly what happens in real tumors, where the physical tumor structure shields inner cells from full drug exposure.
The core finding of this study, that a 3D tumor model more accurately captures drug resistance than 2D culture, has important implications for how cancer drugs are tested before reaching patients. If drugs are evaluated only in flat dishes, they will appear more effective than they truly are, leading to drugs advancing into clinical trials that are likely to fail.
By using models like this GelMA platform, researchers can screen drug candidates more rigorously before committing to expensive and time-consuming human trials. This could accelerate the identification of truly effective treatments and reduce patient exposure to ineffective therapies. The model is also aligned with the scientific principle of the 3Rs (Replace, Reduce, Refine) in preclinical research, providing a human-relevant alternative that reduces reliance on animal experiments.
The authors note that future improvements could include additional components of the tumor microenvironment, such as immune T cells, patient-derived cancer cells (organoids), or drug-resistant cell variants. Combining this 3D platform with high-throughput screening technology could eventually allow personalized drug testing, where a mini-tumor built from a patient's own cells is used to predict which drugs will work best for that individual.
The GelMA hydrogels were extensively characterized to confirm their suitability as a tumor model scaffold. Physical testing showed the gel had a porous, interconnected microstructure visible by electron microscopy, allowing nutrient and oxygen diffusion throughout the matrix. The elastic storage modulus (stiffness) of 5% GelMA matched the mechanical properties of soft biological tissue, which is important because cells respond differently to stiff versus soft environments.
The gel demonstrated controllable swelling and enzymatic degradation, meaning it absorbs water to a stable equilibrium and can be broken down by enzymes present in tissue. This allows the gel to degrade at a physiologically relevant rate rather than persisting indefinitely, as synthetic plastic scaffolds would. Degradation products from gelatin-based materials are non-toxic, making GelMA safer for cell culture than many synthetic polymers.
Infrared spectroscopy and nuclear magnetic resonance confirmed successful methacrylation of gelatin (the chemical modification that allows UV crosslinking), and the final crosslinked gels retained chemical properties suitable for cell adhesion and growth. Together, these characteristics make 5% GelMA a well-characterized and reproducible platform for 3D CRC modeling.