Cas12a2 targeting cancer cells by turning mutant RNA into a trigger

Cas12a2 targeting cancer cells works very differently here from conventional CRISPR genome editing. Instead of correcting a mutation, the system uses mutant RNA as a trigger: once Cas12a2 recognizes the target transcript, it becomes activated and begins damaging chromatin, the DNA packaged inside the nucleus. The resulting genomic damage can stop cell division and lead to cell death.

That logic is particularly attractive for cancer mutations that are difficult to attack with conventional drugs. TP53 is a major example. Its protein product, p53, is altered in roughly 40–50% of cancers, yet many mutant forms lack an obvious pocket for a small-molecule drug. Rather than trying to repair or directly inhibit the mutant protein, the researchers used its RNA transcript as an identifying feature of the cancer cell.

Turning mutant RNA into a trigger

Cas12a2 belongs to the CRISPR family but behaves differently from the better-known Cas9 enzyme. A guide RNA directs it toward a specific RNA sequence. When that target is recognized, Cas12a2 switches into an activated state and gains broad cleavage activity against other nucleic acids nearby.

In bacteria, this activity forms part of an antiviral defence mechanism. Zeng and colleagues repurposed it for a very different purpose: not to make a precise genomic edit, but to unleash destructive nuclease activity inside a selected mammalian cell.

Biochemical experiments showed that activated Cas12a2 could attack chromatin as well as free DNA. In mammalian cells, RNA-triggered activation produced strong DNA-damage signals, activated a genotoxic stress response, halted the cell cycle and promoted cell death.

The central challenge is therefore selectivity. The destructive activity needs to turn on only in cells carrying the intended RNA signature.

One of the clearest tests involved PC9 non-small-cell lung cancer cells carrying the EGFR E746_A750 deletion. This mutation removes 15 nucleotides and creates a new junction in the resulting RNA that is absent from wild-type EGFR.

The team designed a guide spanning that mutant junction. Cas12a2 activation strongly impaired the growth of cells carrying the deletion and produced DNA-damage signalling, while cells with wild-type EGFR were substantially less affected.

Distinguishing a single altered nucleotide

A more demanding problem is distinguishing two RNA molecules that differ by only one nucleotide. The researchers tested several TP53 point mutations, including R248Q and R280K. For some variants, guide RNAs could be designed so that mutant transcripts activated Cas12a2 much more efficiently than their normal counterparts.

The effect was not limited to artificially introduced targets. PC9 cells naturally carry the TP53 R248Q mutation. Directing Cas12a2 against this endogenous transcript impaired cell growth, increased DNA-damage signalling and raised the proportion of dying cells.

The system has important molecular constraints. Cas12a2 cannot recognize every sequence equally well because efficient activation depends partly on a neighbouring RNA feature known as a protospacer flanking site, or PFS. Target transcript abundance also matters. A cancer mutation is therefore not automatically suitable for this strategy simply because it is tumor-specific.

What happened in mice

For the in vivo experiments, the researchers packaged messenger RNA encoding Cas12a2 together with its guide RNA into lipid nanoparticles.

The strategy was tested in several preclinical tumor models. In a MYC-driven liver cancer model, treatment with Cas12a2 directed against MYC RNA reduced measures of liver tumor burden compared with control treatments.

Another series of experiments used PC9 cells to establish lung tumors in immunodeficient mice. In an early-stage model, treatment began five days after 200,000 tumor cells were introduced. Six nanoparticle doses carrying a guide against TP53 R248Q slowed tumor progression relative to control groups.

A more advanced model exposed an important limitation. When one million tumor cells were introduced and treatment was delayed for 21 days, targeting TP53 R248Q no longer reduced the existing tumor burden in the lungs, although the appearance of metastatic disease was delayed.

Delivery also remains a major obstacle. In a separate reporter experiment, a single dose of the lipid nanoparticles delivered RNA to only about 7–18% of PC9 tumor cells in the lungs.

This is therefore not a demonstrated treatment for human cancer. The study is a preclinical proof of principle using cultured cells and mouse models. Human translation would require much more effective delivery, reliable discrimination between mutant and normal transcripts, and careful testing of what happens if an activated chromatin-damaging nuclease reaches the wrong cell.

The conceptual shift is nevertheless important. If a mutant protein is difficult to inhibit and difficult to restore, its RNA can potentially serve as the address of the cancer cell. The mutation then becomes more than a driver or marker of disease: it becomes the signal that activates destructive CRISPR activity inside the cell carrying it.