CTA: How are these platforms being used for cancer research?
Dr Carroll: Quite a number of groups are using whole-genome libraries of guide RNAs to discover potential new cancer drug targets [with CRISPR-Cas9]. The idea is to knock out each gene individually and to identify those that particular cancer cells rely on for their growth. Such experiments are beginning to define regulatory pathways that are essential for cancer cell growth. With luck, this will lead to the discovery of small molecule inhibitors of those pathways that can be developed into drug treatments.
Continue Reading
CTA: What about applications in cancer treatment?
Dr Carroll: Using CRISPR-Cas9 directly for cancer treatment is challenging. One might imagine trying to inactivate a tumor-driver gene, like activated Ras, by targeting it with specific guide RNAs. The problem is that CRISPR knockout is never 100% efficient, and any remaining unmodified cells will rapidly take over again.
CTA: On June 21, a US National Institute of Health (NIH) advisory committee approved a proposal to use CRISPR-Cas9 with chimeric antigen receptor (CAR) T-cell immunotherapy in a human cancer clinical trial. Other, similar trials have been proposed, such as a lung cancer study in China. How will these proposed studies employ CRISPR-Cas9?
Dr Carroll: The particular human trials that are being initiated use CRISPR-Cas9 to knock out genes in therapeutic T cells, like PD-1 in the Chinese trial. The French company Cellectis has modified CAR T cells directed at a leukemia antigen by inactivating genes that contribute to rejection. Thus, the focus is on enhancing the effectiveness of cell therapies, rather than targeting cancer cells directly with CRISPR-Cas9.1,2
CTA: How important are off-target effects and unintended DNA repair effects? (Why do these occur? Are there efforts under way to circumvent these problems with modifications to CRISPR-Cas9 or other genome-editing platforms?)
Dr Carroll: The principal type of off-target effect that concerns people is nuclease cleavage at genomic sequences other than the 1 intended. This occurs because the nucleases are not perfectly specific. For example, Cas9 will bind to and make a break at sequences that have 1 or a few mismatches with the guide RNA.
Because the human genome is big—3 billion base pairs—there are often quite a number of genomic sequences that are related to the target. Cleavage of those secondary targets leads to local mutagenesis, and if those sites are in places where they could alter cell growth properties, they could lead to problems. On one hand, researchers have developed very sensitive methods to detect where those secondary targets are in the genome, so we can think about whether mutations there might be deleterious. In addition, alterations have been made in the Cas9 protein and in guide RNA design that improve the discrimination against related sequences.
Ultimately, we won’t know whether these off-target effects are significant until particular therapies are tried. At this point, I would say the situation looks pretty good when the best designs are used.
