New single-cell technology points to mechanism of cancer immunotherapy resistance

Researchers used Perturb-CITE-seq to identify a protein involved in resistance to immune checkpoint inhibitors.

A 3D structure of a melanoma cell derived by ion abrasion scanning electron microscopy.
Credit: Sriram Subramaniam, National Cancer Institute, National Institutes of Health
A 3D structure of a melanoma cell derived by ion abrasion scanning electron microscopy.

Cancer drugs called immune checkpoint inhibitors (ICIs) stimulate the immune system to kill cancer cells, but many patients develop resistance to these inhibitors. Researchers from the Broad Institute of MIT and Harvard, Columbia University Medical Center, and other institutions have identified a molecular mechanism underlying this drug resistance in melanoma.

The team made this discovery using a technology they developed called Perturb-CITE-seq, which allows researchers to systematically perturb many individual genes and profile RNA and protein in single cells. The combination of a genetic perturbation technology (Perturb-seq) with a method for RNA- and protein-profiling (CITE-seq) lets scientists probe the function of genes, by learning how the activity of a particular gene affects the expression of other genes and proteins throughout the cell. 

In a Nature Genetics study, a team led by Broad core institute member Aviv Regev (on leave, now at Genentech) and Columbia University professor Benjamin Izar describe Perturb-CITE-seq and show how they used it to identify a protein called CD58 as an important mediator of ICI response in human melanoma cells. The findings suggest possible targets for drugs that could one day benefit patients who are resistant to current ICI treatments.

Tumor + T cells

In the study, the researchers grew melanoma cells and tumor-attacking T cells from the same patient together in the same dish. They then used Perturb-CITE-seq to introduce alterations in 750 different genes in the melanoma cells. By analyzing RNA levels and expression of 20 key proteins in both the melanoma cells, the team observed how each gene perturbation affected the relationship between cancer cells and T cells. The scientists collected these data for 218,000 cells and also developed a computational model to analyze the data on a single-cell level.

“What is especially informative with this model is the ability to ‘treat’ a patient’s tumor with their own T cells in a laboratory setting – the same way a researcher might screen drugs in a cell line,” said Broad graduate student and study co-first author Chris Frangieh.

In addition to confirming known genes involved in ICI response, the researchers found that melanoma cells that produced no or decreased levels of the CD58 protein resisted attack from the T cells. This protein hasn’t been found in studies that used animal models, likely because no CD58 counterpart exists in mice. “It really highlights the importance of performing these screens in human models,” said co-first author Johannes Melms, a postdoctoral fellow at Columbia University.

In a follow-up experiment, the researchers deleted CD58 in melanoma cell lines and confirmed that this increased the cells’ survival against T-cell mediated killing. The findings suggest CD58 plays an important role in enabling melanoma to evade the immune system, and may inform the development of new therapies that boost the immune system’s ability to kill melanoma cells.

The authors hope researchers will use Perturb-CITE-seq to better understand other diseases. “Adding a protein read-out expands the types of biological questions we can ask with our screens,” said Broad research fellow and co-first author Pratiksha Thakore.

“We are excited to see what other researchers and clinicians will discover with this technology,” added Katie Geiger-Schuller, co-first author and a Broad research fellow.

The research was funded by the Koch Institute-Dana-Farber/Harvard Cancer Center Bridge Project Grant, the Klarman Cell Observatory, Howard Hughes Medical Institute, the NHGRI Center of Excellence in Genome Science (CEGS), the Center for Cell Circuits, the National Cancer Institute, the Burroughs Wellcome Fund Career Award for Medical Scientists, the Louis V. Gerstner, Jr. Scholars Program, the Velocity Fellow Program, and the National Institutes of Health. 

Paper(s) cited

Frangieh, CJ, Melms, JC, Thakore, PI, Geiger-Schuller, KR. et al. Multimodal pooled Perturb-CITE-seq screens in patient models define mechanisms of cancer immune evasion. Nature Genetics. Online March 1, 2021. DOI: 10.1038/s41588-021-00779-1