Challenges in three-dimensional cell culture: validation of cell viability assays in three-dimensional brain tumour model

Background: Preclinical assays have largely failed to effectively predict the efficacy of anticancer drugs. This could be caused by lack of tumour extracellular matrix components in two-dimensional cell culture. Moving to three-dimensional models, one of the challenges is how to assess drug efficacy. Cell viability is an essential parameter to measure cell growth and to assess the drug effectiveness. Viability assays are typically used in low-cell-density two-dimensional cultures and may not produce the same results in three-dimensional cultures with high cell density and added matrix components.

Aim: Our aim is to validate the common viability assays in three different models: conventional two-dimensional culture, Matrigel-based three-dimensional culture and an organ-specific three-dimensional brain tumour model recently developed by us.

Methods: Cell populations of human glioma U251 cells were plated two-dimensionally in various cell numbers. Identical cell numbers were mixed with Matrigel to form a three-dimensional culture model. Organ-specific three-dimensional scaffolds were developed by decellularizing animal brain tissue followed by seeding with U251 cells. Scaffolds were cultured for up to 3 weeks. We tested the performance of four different viability assays (MTT, Alamar Blue, ATP and LDH) on three different culture models and calibrated the assay output versus known cell numbers by counting and DNA quantification.

Results: In all of our selected cell viability assays, the assay output was higher in two-dimensional samples compared with Matrigel samples containing the same number of cells. In MTT assay by increasing the reagent concentration, the difference between assay reading in two-dimensional and Matrigel samples became less, which might indicate that there is a decreased reagent availability to the cells grown in Matrigel.

Conclusions: Most protocols for viability assays that are validated for two-dimensional cell culture are not (directly) applicable to three-dimensional models; however, optimization and calibration can make them relevant for three-dimensional cell culture.