Doudna lab finds chromosome loss from CRISPR-Cas9 is common—and swapping steps is key to stopping it

When it comes to solving math problems, the order of operations matters (PEMDAS, from left to right!). That appears to be the case for engineering T cells with CRISPR-Cas9, too, as new research from the lab of the technology’s co-creator Jennifer Doudna, Ph.D., shows.

In a paper (PDF) published Oct. 3 in Cell, a research team led by Doudna's lab at the University of California, Berkeley and researcher Chun Jimmie Ye, Ph.D., from the University of California, San Francisco—with help from the labs of CAR-T cell pioneer Carl June, M.D., and Howard Chang, M.D., Ph.D., at Stanford University—described how it discovered that the order of steps performed in CRISPR-Cas9 gene editing changes how well T cells are protected from a phenomenon known as chromosome loss.

“We’re really excited about the fact that not only does our study elucidate fundamental answers about this unintended consequence of CRISPR-Cas9 genome editing, but it also reveals a method for avoiding this potential genotoxicity, which hasn’t been demonstrated before, for current and future trials engineering T cells,” first author Connor Tsuchida, Ph.D., a bioengineer in the Doudna lab, said in a press release.

CRISPR-Cas9 is gaining traction as a method for editing T cells for immunotherapies on account of its precision and efficacy. But some researchers have raised concerns about the possibility of what’s known as aneuploidy, or an abnormal number of chromosomes, in some T cells that are engineered with CRISPR. Specifically, they appear to lose chromosomes, which ultimately leads to cell death, or apoptosis.  

To get a better idea of just how pervasive this problem is, the researchers used CRISPR-Cas9 to edit 92 genes on 22 chromosomes across separate groups of human T cells, focusing on common clinical targets for T-cell engineering and gene therapy. They tested 364 corresponding guide RNAs, or gRNAs—a short sequence of RNA that guides the Cas9 enzyme to the DNA it’s supposed to edit—to see how the type of gRNA used impacted chromosome loss.

Findings from single-cell RNA sequencing analysis of the T cells suggested the phenomenon was ubiquitous. While some chromosome loss was expected as part of the editing process, the researchers found that it occurred above background levels with 55% of the gRNAs used and for 89% of the genes targeted. All of the chromosomes were affected; some were lost completely.

“Chromosome loss was much higher in chromosomes targeted by Cas9 compared with non-targeted chromosomes, suggesting that this phenomenon is an outcome of target-specific cleavage during Cas9-mediated genome editing,” the researchers wrote in the paper.

Additional tests showed that T cells with chromosome loss also had biomarkers of DNA damage and apoptosis. When the scientists monitored cultures of edited T cells over two to three weeks, they saw that the number of them with chromosome loss declined over time, indicating that they were less likely to survive. And while some T cells that had lost chromosomes were able to proliferate, they did so at a lower rate.

Next, the researchers wanted to see whether they could find evidence of chromosome loss in CAR-T cells, which are created by integrating a transgene for a cancer cell antigen receptor into the cells’ DNA. After using CRISPR-Cas9 to insert the transgene into chromosome 14 of human T cells, they again used single cell RNA sequencing to analyze them. A pattern similar to the one they observed in the other engineered T cells emerged—chromosome 14 was lost too frequently to be credited to chance alone.

“Together, these data suggest that chromosome loss is a general phenomenon that occurs in Cas9 edited T cells,” the scientists wrote. “Our findings also indicate that cells with chromosome loss are present among preclinical, Cas9 edited CAR-T cells, highlighting the importance of understanding and mitigating this potential genotoxicity in the context of engineered T cell therapies.”

The researchers sought to understand whether their findings translated to cells that had already been infused into patients. For this, they leveraged a clinical trial for multiple myeloma already underway at the University of Pennsylvania. They collected T cells from two patients and edited them with CRISPR-Cas9, simultaneously targeting three different genes. They then ran single-cell RNA sequencing analysis on the cells before infusing them back into the patients.

The team then collected T cells from the patients at different points in the clinical trial to see how their chromosomes were faring. To their surprise, they observed less than a 1% rate of chromosome loss in the cells, even before they were infused. 

What changed? As it turned out, the researchers had done things differently in the T cells taken from the patients than they had at the bench. In the lab studies, before the T cells were edited, they were activated and stimulated—a key step to ensuring they can target cancer cells once they’re back in the patient’s body. This is the way it’s done in nearly all clinical trials of genome-edited T cells, too, the researchers noted, in order to optimize editing efficiency.

But in the clinical trial at Penn, the scientists swapped the two steps, editing the cells first and stimulating them second. To validate that the different order of operations was protecting the T cells from chromosome loss, the scientists ran an experiment where they edited a common CAR-T cell gene target called TRAC with each of the two methods. This time, they saw a markedly lower level of chromosome loss in the protocol where they edited the cells first—which they dubbed the non-activated T cell editing protocol—than in the one where the cells were activated first.

“Implementation of this modified protocol for Cas9 genome editing in T cells represents a simple adjustment that could substantially mitigate chromosome loss in future research and clinical studies,” the scientists wrote in the paper.