If you have an ACS member number, please enter it here so we can link this account to your membership. (optional)

ACS values your privacy. By submitting your information, you are gaining access to C&EN and subscribing to our weekly newsletter. We use the information you provide to make your reading experience better, and we will never sell your data to third party members.


Biological Chemistry

RNA Interference: the How and Why

Scientists work to figure out the mechanisms underpinning gene-silencing technique

July 5, 2004 | A version of this story appeared in Volume 82, Issue 27


In recent years, scientists have found that RNA--once thought to be little more than a bit player in the pathway from gene to protein--stands more center stage, playing a larger and more varied biological role than previously thought. One of these roles is in the process known as RNA interference, or RNAi. RNAi has generated much excitement within the life sciences community because it allows biologists to do genetic studies that were previously difficult.

Biologists have a much better understanding of the final result of RNAi than they do of how the process itself works. They know in "a big picture sort of way" that RNAi does indeed result in reduced gene expression, but they have only just begun to understand the phenomena at a mechanistic level.

In RNAi, pieces of double-stranded RNA that are 21 to 23 nucleotides long, known as short interfering RNA (siRNA), are incorporated into a protein complex called the RNA-induced silencing complex (RISC). The complex cleaves messenger RNA that is complementary to the siRNA, resulting in reduced expression of a target gene. In essence, RNAi allows biologists to "knock down" gene expression to see how the absence of that gene affects the organism without having to breed "knockouts" that lack the gene.

Among the puzzles surrounding RNAi is that only one of the strands from the RNA duplex is incorporated into RISC. What characteristics determine which strand is picked? For an siRNA to be functional against a target mRNA, the antisense strand must be the one that makes it into RISC, and even then, there's no guarantee it will actually silence the gene. Scientists would like to know how they could ensure that the antisense strand is the chosen one. (The antisense strand is the noncoding strand that is complementary to the strand coding for a gene.)

For biologists who are interested in using RNAi as a tool to study genetics in organisms where such direct studies used to be difficult, if not impossible, it's enough just to know that RNAi works. Other scientists are interested in understanding the process itself--how and why it works--so that they can improve it.

Two such scientists are Phillip D. Zamore, associate professor of biochemistry and molecular pharmacology at the University of Massachusetts Medical School, in Worcester, and Anastasia Khvorova, vice president of research at Dharmacon in Lafayette, Colo. Zamore and Khvorova (in the group led by Sumedha Jayasena at Amgen) independently took different approaches to arrive at the same conclusion: The stability (both absolute and relative) at the 5´ ends of the two siRNA strands determines which one enters RISC [Cell, 115, 199 and 209 (2003)]. The strand with the less stable 5´ end is more likely to be incorporated in RISC because it is easier to unwind. The researchers both presented their work at a meeting on the cell biology of RNAi held at the National Academy of Sciences in May.

A mature RNA-induced silencing complex incorporating the correct short interfering RNA (siRNA) strand recognizes, cleaves, and releases its target mRNA before going back to do it again. The 3´ (hydroxyl) and 5´ (phosphate) ends of the siRNA are indicated.
A mature RNA-induced silencing complex incorporating the correct short interfering RNA (siRNA) strand recognizes, cleaves, and releases its target mRNA before going back to do it again. The 3´ (hydroxyl) and 5´ (phosphate) ends of the siRNA are indicated.

"IT'S BECOMING increasingly clear from our work and complementary work by several other labs that the single most common defect in siRNAs [occurs when] the sense strand ends up in RISC," causing the siRNA to be considered nonfunctional, Zamore explains. "In fact, those types of siRNAs are perfectly functional. It's just that they're not functional against the desired target. My hope is that, with these new insights into the machinery itself, people will be able to avoid that pitfall."

One way to tell which strand will be incorporated into RISC is to look at the "internal stability profile" for the strands, Khvorova says. The internal stability is a thermodynamic measurement that describes how tightly the two RNA strands are held together. Rather than being a single value for the entire duplex, it varies along the length of the strand. It is a property of each individual strand but only in the context of the duplex.

"We realized that when you look at internal stability profiles of functional versus nonfunctional siRNAs, they are drastically different," Khvorova says. The main differences are at the 5´ ends of the strands and the base pairs at positions 11–14, with the antisense strand being less stable in functional siRNAs.

Dharmacon chemically modifies the sense strands of siRNAs to block their entry into RISC. In about 30% of the cases, they were able to substantially improve the functionality of the siRNAs by modifying the sense strands, Khvorova says. She won't identify the exact nature of the modifications, but says that they can be on the backbone, the ribose, or the base.

It turns out, however, that inefficient entry of the antisense strand into RISC is only part of the problem. "It appears that siRNA activity on average is limited by inefficient RISC entry in only 30% of cases," Khvorova says. "You can take 100 nonfunctional siRNAs. You alter the thermodynamics either by chemical modifications or structural elements, and you fully recover the activity of only 30 of those siRNAs. Seventy are still not fully active. You will get efficient entry into RISC, but they are not functional because they are not able to work in the later stages [of the RNAi pathway]." The researchers don't yet know what other aspects of siRNAs affect their functionality.

Zamore uses the term "functional asymmetry" to describe the tendency of one strand of the siRNA to be preferentially incorporated into RISC. "We can take any sequence, including those that start out highly symmetric," meaning that the strands have a close-to-equal likelihood of being incorporated into RISC, "and convert it to one that's functionally asymmetric," he says.

Zamore's strategy for introducing functional asymmetry turns out to be quite simple. "We simply change a single base pair on the sense strand. We usually change position 19 [counting from the 5´ end] so that it's now a mismatch with the first position of the antisense strand." (Although both strands are 21 nucleotides long, they each have a two-nucleotide overhang, so the strands share only 19 base pairs.) "It doesn't seem to matter what the mismatch is, just that it's a mismatch," he says. This mismatch helps destabilize the 5´ end of the antisense strand, making that end easier to unwind. He says that by using this approach, it's even possible to resurrect "dead" siRNAs that incorporate the sense strand or neither strand.

Now that they think they understand the factors that promote efficient incorporation in RISC, Zamore and his colleagues are moving on to try to develop a mechanistic understanding of other aspects of the RNAi pathway. "Part of the groundwork for understanding [the RNAi mechanism] is being able to describe the RNAi pathway in terms of classical enzymology," he says.

He and his colleagues recently published a kinetic analysis of the RNAi enzyme complex [Nat. Struct. Mol. Biol., 11, 599 (2004)]. The analysis suggests that the different regions of the siRNA play distinct roles in the catalytic cycle. The 3´ end of the siRNA seems not to contribute to the binding of the complex to the target mRNA. However, it's also clear that 3´ complementarity is important for siRNA function. "When there's mismatch at the 3´ end of the siRNA with its target, cleavage is slower," Zamore says. "It was surprising to me how cleanly one could separate contributions to binding from contributions to catalytic function."

There is also evidence that siRNA is involved in other biological processes such as the formation of heterochromatin, the DNA packing material found in parts of the chromosomes where there are few genes. An understanding of how siRNA and RNAi work will help illuminate those areas of biology as well.

"It's already clear that siRNA has very deep biological tentacles," Zamore says. "The relationship of the mechanism to the biology is a very interesting problem."


This article has been sent to the following recipient:

Chemistry matters. Join us to get the news you need.