Painting profiles of perturbation to perceive genes’ purpose
A cellular imaging technique called Cell Painting creates visual “signatures” that help reveal the functions of uncharacterized genes or disease-associated gene variants.
By Tom Ulrich
Credit: Lauren Solomon, Broad Communications, and Broad Target Accelerator Team
Scientists have used cells’ visual appearance, or morphology, as a way to help understand their state and identity essentially since the invention of the microscope. Now a research team led by Broad Institute Imaging Platform director Anne Carpenter and postdoctoral fellow Mohammad Rohban has shown that a high-throughput, computerized imaging technique for studying morphology, called Cell Painting, can provide insight into the cellular roles of genes or disease-linked gene alleles whose function or impact is unknown.
With Cell Painting, a techniquedeveloped at the Broad, researchers tag eight cellular components and organelles (actin, cytoplasmic RNA, endoplasmic reticulum, Golgi apparatus, mitochondria, nucleus, nucleolus, and the plasma membrane) with fluorescent dyes. They then image the cells microscopically and computationally generate morphological profiles or signatures — measurable and reproducible changes in more than 1,500 features of the cells’ appearance — that arise in response to a given perturbation.
With Cell Painting, researchers stain cells for several structures and components simultaneously, such as nuclei. (Photo: Broad Target Accelerator Team; click dots below to advance)
The same cells, stained to reveal their nucleoli and endoplasmic reticulum. (Photo: Broad Target Accelerator Team)
And now stained for cytoplasmic RNA. (Photo: Broad Target Accelerator Team)
A merged image of all of the stains used. (Photo: Broad Target Accelerator Team)
In the current study, published online in eLife, Rohban, Carpenter, and their colleagues overexpressed 220 genes one-by-one in cultured cells, dyed the cells, and searched for signatures associated with each gene.
The team found that they could identify such signatures for half of the genes studied. Moreover, they found that they could group genes by function based on those signatures. For instance, overexpressing genes that reshape the cytoskeleton (the cell’s structural scaffolding) resulted in similar signatures, as did overexpressing genes within or related to particular cell signaling pathways.
The findings suggest that Cell Painting may provide researchers with an inexpensive, high-throughput means to understanding the functions of uncharacterized genes — a category that currently includes more than 30 percent of the human genome. In addition, Cell Painting could help identify the biological mechanisms behind genetic variations flagged by genome-wide association and other large-scale genetic studies of traits and disease.