T H E   N I H    C A T A L Y S T      M A R C H  –  A P R I L   2003

Hot Methods

text and photos
by Masashi Rotte


from the Therapeutic Oligonucleotide Interest Group website

Running interference—–that is, interfering with the expression of targeted genes—has become a popular sport in some scientific circles.

Called RNA interference (RNAi), this approach to intervening in biological processes is gathering momentum as researchers refine their techniques and envision widening applications in human health.

RNAi has been the subject of an increasing number of research articles and in February was the theme of two NIH seminars—as well as an e-mail posted on the NIH fellows’ ListServe by an NIMH researcher asking, "Is anybody out there synthesizing siRNAs? We need a considerable amount to study functional roles of a gene possibly involved in mood regulation." [siRNAs are short interfering RNAs used in RNAi in mammals, discussed later.]


RNAi is a naturally occurring form of post-transcriptional gene silencing (PTGS) mediated by double-stranded RNA (dsRNA).

By taking advantage of RNAi mechanisms, researchers have been able to reduce the expression of specific target genes on an individual cell to organism-wide scale.

History: Plants

In the early 1990s, researchers were attempting to enhance the purple color of petunias by introducing a pigment-producing transgene under the control of strong promoter.

Instead of overexpressing the pigment gene and enriching the color of the plants, the flowers appeared nearly white—the expression of both the transgene and the endogenous pigment gene was "co-suppressed." This co-suppression refers to silencing of gene expression principally at a post-transcriptional level.

Worms and Flies

The phenomenon of PTGS was observed in other plants and fungi during the 1990s, but not until work done with the nematode Caenorhabditis elegans by Andrew Fire and his colleagues was a similar effect seen in animals. It was shown that the introduction into embryos of antisense (but, unexpectedly, also sense) dsRNA induced a sequence-specific gene silencing at a post-transcriptional level, an effect that was termed RNAi.

Subsequent studies—one of the first carried out by Leonie Misquitta and Bruce Paterson, NCI—showed a similar response in Drosophila ("Targeted disruption of gene function in Drosophila by RNA interference [RNA-i]: A role for nautilus in embryonic somatic muscle formation," Proc. Natl. Acad. Sci. U S A 96:1451–1456, 1999).

Tips on the RNAiceberg: NHGRI’s Natasha Caplen discussed the "Characterization and Application of RNAi in Mammalian Cells"


Having found RNAi gene silencing in worms and flies, investigators now turned their attention to determining whether RNAi could be induced in mammalian cells. It was known, however, that in somatic mammalian cells, the introduction of long dsRNAs (~70+ nucleotides) led to nonspecific suppression of gene expression instead of the sequence-specific gene silencing seen in RNAi. Studies revealed that long dsRNAs activated either the protein kinase PKR, leading to the repression of translation, or RNaseL, leading to nonspecific RNA degradation. The subsequent global changes in gene expression usually resulted in cell death.

The development of a strategy to overcome this problem was aided by work in plants and subsequently in Drosophila embryos. This work showed that dsRNA added to the cells was processed by an RNase III enzyme called Dicer to nucleotides 21–23 base pairs in length.

Size Matters

These RNAs were termed short interfering RNAs (siRNAs). Their importance in RNAi emerged with the finding that homologous Drosophila mRNA is cleaved at a site corresponding to the approximate middle of an siRNA sequence.

This knowledge, coupled with previous data showing that RNA duplexes of less than 60–70 nucleotides do not trigger PKR, led researchers to test chemically synthesized siRNAs, 21–27 nucleotides long, in mammalian cells.

These landmark studies, including one by NHGRI’s Natasha Caplen, showed that synthetic siRNAs of 21–23 nucleotides in length introduced by transient transfection effectively induce RNAi in mammalian cultured cells in a sequence-specific manner (N. Caplen, S. Parrish, F. Imani, A. Fire, and R.A. Morgan, "Specific inhibition of gene expression by small double-stranded RNAs in invertebrate and vertebrate systems," Proc. Natl. Acad. Sci. U S A 98: 9742–9747, 2001).


The working model that has now emerged for RNAi is as follows:

siRNAs are incorporated into a multiprotein RNA-induced silencing complex, or RISC.

The siRNA within the RISC unwinds, and the antisense strand acts to guide the complex to homologous mRNA transcript by base pairing.

The target mRNA is then catalytically cleaved by an undefined component or components of RISC approximately 12 nucleotides from the 3' terminus of the siRNA.

Recently, another RNA species has also been found to use the enzyme Dicer. First identified in C. elegans, small temporal RNAs (stRNAs) regulate developmental timing through translational repression of target transcripts using Dicer.

stRNAs have now been found to be members of a large family of noncoding RNAs called microRNAs (miRNA), which have been identified in plants, Drosophila, C. elegans, mouse, rat, and human cells. miRNAs require Dicer for their processing and are now believed to play a critical role in regulation of gene expression.

These and other links to endogenous cellular pathways have led to speculation that PTGS evolved as a defense mechanism against transposons and RNA viruses or as a means for the cell to rid itself of unnecessary transcription products.


There are several key considerations when designing siRNAs and implementing RNAi. Two recent seminars on campus—one by Caplen, a staff scientist in the Medical Genetics Branch, NHGRI, and the other by Stephen Scaringe, chief scientific officer of Dharmacon, Inc., Lafayette, Ohio—addressed some of those considerations, as well as recent advances in RNAi.

Taking the Measure of Silence

The efficacy of gene silencing by siRNAs can be evaluated by the drop in mRNA concentration of the gene of interest, by the drop in protein expression, or by functional changes in cell phenotype.

During her talk at the Therapeutic Oligonucleotide Interest Group seminar on February 27, Caplen described the use of GFP-proteins to test silencing efficiency of siRNA. A drop in cell fluorescence measured by flow cytometry corresponds to an RNAi-mediated silencing of gene expression.

The most effective siRNAs can reduce mRNA levels and expression of protein by greater than 90 percent. A primary factor that determines the efficacy of silencing is the selection of optimal siRNA sequences to the target mRNA: siRNAs that target different sequences within the same mRNA can show varying degrees of efficacy.

Maximizing Beginners’ Luck: Stephen Scaringe, of Dharmacon, Inc., offered best practices in siRNA design, target selection for optimal knockdown, and pooling to maximize the chance of success the "first time out"

Designer Basics

Caplen gave some of the basic criteria for design of siRNAs:

They must be 21–23 nucleotides long.

They should have sequence-specific homology to the target mRNA.

They should have as close to a 50 percent GC content as possible.

They should have two nucleotide 3' overhangs.

Researchers are advised to synthesize siRNAs to several sequences in the target mRNA (usually two or three). Having identified a successful siRNA, researchers may then want to make base pairs iterations upstream or downstream of the initial sequence as a way to optimize their siRNA.

siRNA Pools

In his talk on February 28 on "Developments in siRNA-based Gene Silencing," Scaringe described software that Dharmacon developed to select the most favorable targets in mRNA sequences based on 34 criteria. (A free software module on his company’s website selects target sequences but uses only four criteria.)

In addition, Dharmacon uses another five criteria to "pool" the four or five most effective siRNAs. Scaringe claims that his company’s algorithms can design siRNA pools that will knock down mRNA levels at least 50 percent. Synthetic siRNA can show RNAi effects six hours post-introduction and the effects can last as long as 10 days, although the degree of silencing will be diminished by that time.

Currently, three main companies offer ready-made, chemically synthesized siRNAs: Dharmacon, Qiagen (Valencia, Calif.), and Ambion (Austin, Texas). Dharmacon has ready-made siRNA kits for several human and mouse genes and also offers custom siRNA synthesis. Qiagen offers a cancer siRNA Oligo set with siRNAs to 139 human cancer genes. Qiagen’s website has a database to help select siRNA targets for mRNA of several species in the GenBank database.

Transfection Techniques

A variety of methods have been used to introduce siRNA into cells and organisms, ranging from microinjection to soaking cells in dsRNA to feeding C. elegans bacteria engineered to express dsRNA. The most common means to introduce siRNA is a lipid-based transfection reagent, but this method is transient at best.

Many researchers are now using plasmids or viral vectors to continually express siRNAs in transiently and stably transfected mammalian cells (mainly from RNA polymerase III promoters using a short-hairpin structure that is processed intracellularly by Dicer to generate an siRNA).

To test transfection efficacy, Caplen has labeled siRNAs with a fluorescence tag. This allows the researcher to monitor uptake of the siRNA and to enrich for cells that have been successfully transfected with fluorescence-activated cell sorting.


Caplen notes that among recent developments in the field is the generation of transgenic mice from embryonic stem cells harboring an siRNA expressed from a viral vector. The investigators successfully recapitulated features of animals produced by traditional homologous recombination knockout technologies. Scanning the recent scientific literature provides a glimpse of some of the targets of RNAi research. Some investigators have determined the specific effects of protein inhibition arising from silencing their gene of interest by using cDNA microarray analysis. Others have used RNAi to silence gene expression in vivo in mice, including a study in which the silencing of Fas gene expression blocked hepatocyte cell death. And still others have investigated siRNAs as a therapy to inhibit expression of oncogenes or to downregulate expression of CD4 cells to block HIV entry.

Once optimal siRNAs have been designed and introduced into a cell, the experimental possibilities are limitless. Caplen says that her main research goals are adapting RNAi for high-throughput functional analysis and the development of RNAi as a therapeutic.

Disclaimer: Mention of specific products in this article does not constitute an endorsement of those products, nor does it signify that other similar products are less desirable.


Return to Table of Contents