Caption:MIT researchers demonstrated that their RNA sensor could accurately identify cells expressing a mutated version of the p53 gene, which drives cancer development. [Image courtesy of MIT/iStock]
Engineers at MIT designed a new way to trigger cells that could lead to targeted therapies for cancer and other diseases.
Using an RNA sensor for the trigger, the researchers ensure that synthetic genes activate only in specific cells. They demonstrated that the sensor could accurately identify cells expressing a mutated version of the p53 gene, which drives cancer development. Additionally, they found they could turn on a gene encoding a fluorescent protein only within those cells.
With this approach, the engineers feel they can develop sensors to trigger production of cell-killing proteins in cancer cells, sparing healthy cells, too. They see the possibility of developing treatments for other diseases, too, including viral or bacterial infections.
“There’s growing interest in reducing off-target effects for therapeutics,” said James Collins, the Termeer Professor of Medical Engineering and Science in MIT’s Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering. “With this system, we could target very specific disease cells and tissues, which opens up the possibility of identifying cancer cells and then delivering highly potent therapeutics.”
Collins served as senior author of the engineers’ study, which they published in Nature Communications. MIT postdocs Raphaël Gayet PhD ’22 and Katherine Ilia PhD ’23, senior postdoc Shiva Razavi, and former postdoc Nathaniel Tippens served as lead authors.
The control switch
Collins’ lab in 2021 developed the eToehold, a control switch for RNA therapies. They based it on RNA molecules called internal ribosome entry sites (IRES). Researchers can design these to respond to a particular messenger RNA (mRNA) sequence within a cell. According to the team, these systems are difficult to design, though. Their function depends on the sequence of the IRES molecule and its 3D structure.
In their study, the researchers wanted to create a system they could program more easily. They replaced IRESes with a synthetic strand of RNA — called an RNA construct — as the targeting molecule. The team said this allows for reprogramming the construct and targeting different mRNA molecules by changing the RNA sequence.
“With this new system, we have a very straightforward, programmable way of creating control elements that will respond only in the presence of those target sequences,” Collins said.
Allowing the cell to read the RNA construct
The team then harnessed an enzyme existing in most animal cells known as adenosine deaminase acting on RNA (ADAR). It helps cells fight off invading viruses, among other things. ADAR can detect and repair mismatches in double-stranded RNA. The researchers designed the sensor RNA to contain a sequence complementary to the target RNA but with a lone mismatch. It draws the ADAR’s attention, repairing the mismatch.
ADAR converts adenosine to inosine in the RNA sensor, removing a stop codon in the sequence. The cell begins reading the RNA construct, designed to contain two protein-coding genes. One, a fluorescent protein, lets the researchers see the gene activation. They say they could replace this in future versions with a gene encoding a therapeutic agent. The other gene encodes a stripped-down version of the ADAR enzyme. This creates a positive feedback loop that enhances the expression of the fluorescent reporter gene.
“We only require a very small amount of ADAR to initially trigger the network. And then through a positive feedback design, that small trigger gets the cells to express high levels of a compact form of that enzyme that’s built into the construct,” Collins says. “This broadens the potential application uses for the system in that now it’s not restricted to cells that have large background levels of ADAR.”
Utilizing precision
The researchers named their cell DART VADAR (detection and amplification of RNA triggers via ADAR). They evaluated the sensor in human cell tests to see if it could distinguish between very similar mRNA sequences. To do so, they inserted the sensor construct into human cells with either the normal p53 gene or a mutated version.
“We show that you can get very high resolution and very high precision for these sensors,” Gayet said. “With a carefully designed sensor, you can get a different level of activation depending on whether or not the cells produce some RNA that includes a mutation.”
Another set of experiments in mouse cells demonstrated the ability of the sensor to distinguish between closely related cell types. These differentiate into either bone or muscle cells.
By utilizing a trimmed-down version of ADAR, the researchers say the construct fits easily into an AAV vector. This modified, harmless virus often (clinically) delivers genetic material in humans. The researchers now plan to try the system in animal models of cancers. They aim to see if they can deliver synthetic constructs that selectively kill tumor cells by producing lethal compounds within only those cells.
The National Institutes of Health and the Wyss Institute for Biologically Inspired Engineering funded the research.