NEW YORK – A UK-led team used a combination of high-throughput gene editing and transcriptomics to prospectively characterize drug resistance mechanisms in several cancer cell lines, identifying variants that may inform future treatment strategies.
The findings appeared in Nature Genetics on Friday.
"[T]his study could help to understand what variants to look for in biopsy samples and in what proteins, and how to interpret the findings," senior and co-corresponding author Mathew Garnett, a translational cancer genomics researcher affiliated with the Wellcome Sanger Institute and Open Targets, said in an email.
Starting with four cancer cell lines representing lung cancer, colon cancer, and Ewing sarcoma, Garnett and colleagues established a guide RNA library to systematically introduce nearly 32,500 genetic variants into 11 cancer-related genes with CRISPR base editing, tracking the growth and drug response effects of this targeted mutagenesis screening approach.
After exposing the edited cells to 10 targeted oncology drugs, the team used a modified version of the perturb-seq approach focused on the gene expression consequences of edited variants linked to drug sensitivity or resistance. Together, the results highlighted treatment response-related variants falling into four functional groups.
While "driver" variants appeared to boost the growth of cancer cells regardless of drug exposure, a set of "canonical drug resistance" variants appeared to bump up cancer cell proliferation in the presence of drug treatments but not in the drug-free setting. Another group of variants, known as "drug-sensitizing" variants, increased cancer cell vulnerability to specific drugs.
Finally, the team noted that a group of so-called "drug addiction" variants bumped up cancer cell viability and growth in the presence of anti-cancer drugs, despite their negative effects on cancer cell proliferation in the absence of such drugs.
The resulting "variant-to-function" map "has implications for patient stratification, therapy combinations, and drug scheduling in cancer treatment," the authors suggested, adding that the map is expected to "inform the interpretation of cancer genomics data in the clinical management of drug-resistant cancers."
In a lung cancer cell model, for example, the team flagged epidermal growth factor receptor variants that enhanced sensitivity to EGFR inhibitors, highlighting the possibility of identifying patients with pronounced responses to such drugs with tumor testing or genomic profiling. On the other hand, "drug addiction" resistance variants that are detrimental in the absence of a therapeutic challenge inform treatment scheduling.
"The ability to detect the emergence of populations of cells with resistance variants early could help guide second-line or combination therapies earlier," Garnett explained, adding that the presence of a "drug addiction" variant "would predict that a patient could benefit from a break from drug treatment — a so called 'drug holiday.'"
The current findings also point to the possibility of performing similar, but still broader, prospective analyses on drug resistance in the future — from efforts to interrogate cancer cell responses to distinct drugs with shared treatment targets to studies comparing treatment resistance features found in different cancer types.
Moreover, the study's authors noted that "[a]dvancements in base editing and prime editing technology could increase the editing saturation and accuracy and improve the interpretation of negative results, including the identification of benign variants without the need for extensive genotyping."