Cell Painting method reveals drug differences in 3D cancer cells

Researchers at Uppsala University have adapted Cell Painting, a widely used imaging method in drug discovery, to look at individual cells inside small 3D clusters of cancer cells. Two drugs that looked almost identical in ordinary flat cell cultures behaved differently in 3D, and the team linked the difference to how each drug switches on the cell's p53 stress response.

Cell Painting colours the main parts of a cell, such as the nucleus, mitochondria and the cell's internal scaffolding, so that they can be photographed in detail under a microscope. Drug developers use it to see how drugs change cells. Until now it has mostly been used on cells grown as a flat layer, even though cells in the body grow in three dimensions.

The researchers grew bowel cancer cells as small spheres, or minispheres, in standard laboratory plates. They then made the minispheres transparent with a simple solution of sugar and glycerol, so the microscope could see individual cells deep inside them, and measured thousands of details of each cell's shape and structure.

"Cells behave differently when they grow in three dimensions, but most screening methods either use flat cultures or collapse a whole 3D sample into one picture. We wanted to keep the detail of single cells while still being able to test many drugs at once," says first author Christa Ringers, researcher at the Department of Pharmaceutical Biosciences, Uppsala University, who shares first authorship with David Holmberg. 

The team tested 52 drugs used in or developed for cancer treatment. Two of them, the chemotherapy 5-fluorouracil and the targeted cancer drug olaparib, were nearly indistinguishable in flat cultures but clearly different in minispheres. The difference only appeared when the researchers looked at each cell separately, not when they looked at each minisphere as a whole. 

Follow-up experiments revealed a surprise. 5-fluorouracil caused less DNA damage than olaparib, yet it was better at stopping cells from dividing. Gene activity pointed to the p53 pathway, a central stress response in cells, which 5-fluorouracil activated more strongly.

"Together with colleagues at AstraZeneca, we also ran the method outside the lab where it was developed. That kind of collaboration is a valuable test of whether a new method can be used by others," says Ola Spjuth, professor at the Department of Pharmaceutical Biosciences, Uppsala University, who supervised the study together with Jordi Carreras-Puigvert.

The study was done in two cancer cell lines in the laboratory and does not show how the drugs work in patients. The researchers now hope to apply the method to more complex 3D models, such as organoids grown from patient tissue. All images, data and code are openly available.