Researchers develop new RNA molecule approach to kill cancer cells

Sep 8 2010

Researchers at the California Institute of Technology (Caltech) have engineered a fundamentally new approach to killing cancer cells. The process-developed by Niles Pierce, associate professor of applied and computational mathematics and bioengineering at Caltech, and his colleagues-uses small RNA molecules that can be programmed to attack only specific cancer cells; then, by changing shape, those molecules cause the cancer cells to self-destruct.

In conventional chemotherapy treatments for cancer, patients are given drugs that target cell behaviors typical of-but not exclusive to-cancer cells. For example, cancer drugs commonly attack cells that divide rapidly, because such accelerated division is a hallmark of most cancer cells. Unfortunately, rapid cell division is a property of normal cells in the bone marrow, digestive tract, and hair follicles, and so these cells are also killed, leading to a host of debilitating side effects.

A better method, says Pierce, is to create drugs that can first distinguish cancer cells from healthy cells and then, once those cells have been spotted, mark them for destruction; in other words, to produce molecules that diagnose cancer cells before eradicating them. This type of therapy could do away with the side effects associated with conventional chemotherapy treatments. It also could be tailored on a molecular level to individual cancers, making it uniquely specific.

In a paper slated to appear online the week of September 6 in the Proceedings of the National Academy of Sciences (PNAS), Pierce and his colleagues describe just such a process. It employs hairpin-shaped molecules known as small conditional RNAs, which are less than 30 base pairs in length. (An average gene is thousands of base pairs long.)

The researchers' method involves the use of two different varieties of small conditional RNA. One is designed to be complementary to, and thus to bind to, an RNA sequence unique to a particular cancer cell-say, the cells of a glioblastoma, an aggressive brain tumor. In order to bind to that cancer mutation, the RNA hairpin must open-changing the molecule from one form into another-which, in turn, exposes a sequence that can spontaneously bind to the second type of RNA hairpin. The opening of the second hairpin then reveals a sequence that binds to the first type of hairpin, and so on.

In this way, detection of the RNA cancer marker triggers the self-assembly of a long double-stranded RNA polymer. As part of an innate antiviral immune response, human cells defend against infection using a protein called protein kinase R (PKR) to search for long double-stranded viral RNA, which should not be present in healthy human cells. If PKR indeed detects long double-stranded RNA within a cell, the protein triggers a cell-death pathway to eliminate the cell. "The small conditional RNAs trick cancer cells into self-destructing by selectively forming long double-stranded RNA polymers that mimic viral RNA," says Pierce. "There is, however, no virus."

Pierce and his colleagues tested the process on lab-grown human cells derived from three types of cancers: glioblastoma, prostate carcinoma, and Ewing's sarcoma (a type of bone tumor). "We used three different pairs of small conditional RNAs," with each pair designed to recognize a marker found in one of the three types of cancer, he explains. "The molecules caused a 20- to 100-fold drop in the numbers of cancer cells containing the targeted RNA cancer markers, but no measurable reduction in cells lacking the markers." For example, he explains, "drug 1 killed cancer 1 but not cancers 2 and 3, while drug 2 killed cancer 2 but not cancers 1 and 3, and drug 3 killed cancer 3 but not cancers 1 and 2."

"Conceptually," Pierce says, "small conditional RNAs provide a versatile framework for diagnosing and treating disease one cell at a time within the human body. However," he notes, "many years of work remain to establish whether the conceptual promise of small conditional RNAs can be realized in human patients."

Source: California Institute of Technology