Researchers from Penn Medicine and the Children’s Hospital of Philadelphia (CHOP) have developed a new CRISPR-based platform that can identify genes and regulatory elements driving acute myeloid leukemia (AML) directly in patient cells. The study, published in Molecular Cell, represents the first time this approach has been used on patient-derived cancer samples rather than preclinical models or long-established cell lines.
CRISPR genome-editing tools have allowed scientists to test hundreds of genes at once to determine which are essential for cancer growth and survival. Traditionally, these tests were conducted on laboratory-grown cell lines that do not fully capture the genetic diversity seen in patients. By applying CRISPR technology directly to heterogeneous tumor samples from patients, researchers aim to better understand how specific genetic elements influence cancer cell behavior and treatment response.
AML is responsible for about one-third of adult leukemia cases and is the second most common blood cancer among children in the United States. While chemotherapy can induce remission in many AML patients, some experience relapse or resistance due to genetic changes within their leukemia cells. The new tool could eventually help prioritize treatment options based on each patient's unique cancer biology.
The research team optimized viral vectors and delivery methods to introduce CRISPR components into primary leukemia cells from patients, achieving high gene-editing efficiency. Using this system, they screened hundreds of gene edits simultaneously to find those affecting cell growth—either reducing or increasing it—which points to genes important for cancer survival. These findings were validated both in vitro and using transplanted patient-derived leukemia cells in preclinical models.
“This platform empowers scientists to test which genes and genetic elements really matter in human tumors,” said Junwei Shi, PhD, lead author of the study and associate professor of Cancer Biology at the Perelman School of Medicine at the University of Pennsylvania. “It helps identify drug-ready targets, shows how different tumor subpopulations within the same patient respond and speeds discovery of precision therapies.”
Single-gene edits succeeded in about 86 percent of patient samples tested, while high-throughput screening worked in roughly 73 percent. The researchers confirmed several known leukemia "dependency" genes as well as vulnerabilities unique to certain patients or subtypes.
To gain further insight into cellular responses, the team combined CRISPR editing with single-cell RNA sequencing. This method revealed how individual cells changed their gene activity, state, and behavior after specific edits—capturing diverse responses among different cell populations within a single patient's leukemia sample.
“We have learned that most leukemias are heterogeneous and may contain small subgroups of cells that may ultimately drive poor outcomes,” said Kai Tan, PhD, senior study author and professor in CHOP’s department of Pediatrics. “This clarified the results and revealed surprises. We validated previously reported genes that affect leukemia growth but also found that some edits caused cells to die while others halted growth and induced a dormant, therapy resistant state. These insights will help prioritize the best candidate genes for therapy development.”
The team plans to extend this research by studying other hard-to-treat leukemias such as pediatric AML.
“Our hope is that this novel platform will identify new ways of developing precision therapies for patients who do not currently have promising options,” said Kathrin M. Bernt, MD, senior study author and pediatric oncologist at CHOP’s Cancer Center Leukemia and Lymphoma Program.
Funding for this work came from organizations including St. Jude Children’s Research Hospital Collaborative Research Consortium on Novel Therapies for Sickle Cell Disease; Mark Foundation for Cancer Research; FDA; Pediatric High-Risk Cancer Preclinical Model Resource; National Institutes of Health grants U54CA283759, CA226187, CA243072, CA233285, CA201230; and CA258904.
