By Mehdi Banan, Ph.D.
Science Content/Business Development Analyst

Introduction of long double-stranded RNA (dsRNA) into cells derived from certain organisms (such as plants, fungi, fruit flies, and worms) causes degradation of homologous mRNAs--a phenomenon referred to as RNA interference (RNAi). RNAi takes place in two steps. First, the long dsRNAs are cleaved into 21-25 bp small interfering RNAs (siRNAs) by the ribonuclease Dicer. The siRNAs subsequently assemble with protein components into an RNA-induced silencing complex (RISC) that tags and destroys cognate mRNAs. Two years ago, Elbashir and Tuschl used siRNAs to silence specific genes in mammalian cells (1). Since then, siRNAs have been used extensively for this purpose.

siRNA-mediated gene silencing is generally believed to be highly sequence-specific. Tuschl and colleagues demonstrated that even a single base mismatch between an siRNA and its mRNA target abolished gene silencing (2). This does not preclude the existence of non-specific gene silencing by siRNAs, however. It is possible that siRNAs non-specifically silence a subset of partially complementary mRNAs. siRNAs may also induce nonspecific gene silencing through initiation of an antiviral response similar to the one that occurs following introduction of long (>30 bp) dsRNA--a response that is mediated through blockage of protein synthesis (via activation of protein kinase R and the subsequent phosphorylation of EIF2a) and nonspecific RNA degradation (via activation of RNase L).

These concerns necessitate that the extent of siRNA specificity be explored on a genome-wide scale. In the 2 May 2003 issue of The Proceedings of the National Academy of Sciences, USA, Patrick Brown and colleagues examine the genome-wide specificity of siRNAs by means of DNA microarrays (3). Specifically, they showed that siRNAs directed toward an exogenous GFP gene did not affect the expression levels of the ~36,000 human genes that were spotted on the microarray.

The Results

Brown and colleagues examined the specificity of two siRNAs that targeted an exogenous GFP gene, reasoning that a knockdown of GFP would not indirectly alter the expression levels of cellular genes. To this end, two GFP-specific siRNAs (and two scrambled sequence control siRNAs) were transfected separately into human 293 cells that transiently or stably expressed GFP. As predicted, the GFP-specific siRNAs silenced GFP expression by >70%, whereas the control siRNAs had no effect on GFP levels (as determined by GFP fluorescence levels and FACS analysis). The RNA from the siRNA-transfected and non-transfected cells was then isolated, labeled, and hybridized to a cDNA microarray containing ~36,000 human gene sequences. Approximately 20,000 of these genes are expressed in 293 cells. The expression level of none of these 20,000 genes was significantly altered upon introduction of GFP-specific or control siRNAs into cells.

Another concern with RNAi specificity is the generation of secondary siRNAs by "transitive RNAi". Transitive RNAi is a phenomenon that takes place in C. elegans that may also occur in other organisms. In this process, the siRNAs, after being denatured, bind to complementary transcripts and generate new dsRNAs via a 5' to 3' extension reaction carried out by an RNA-dependent RNA polymerase (RdRP). These dsRNAs, in turn, are processed into so-called 'secondary siRNAs'. These secondary siRNAs, if complementary to other gene sequences, could silence the expression of those genes--greatly enhancing the probability of nonspecific gene silencing.

To date, no RdRP has been found in humans, suggesting that transitive RNAi does not take place in human cells. This premise, however, has not been extensively addressed. Brown and colleagues explored the existence of transitive RNAi in human cells. To this end, they co-transfected two reporter gene constructs (luciferase and GFP) with a common fused actin sequence into human 293 cells. They then determined that an siRNA to one reporter gene did not silence the other. These results suggest that a transitive siRNA mechanism does not exist in humans.

Discussion

The above findings support the premise that siRNAs function in a highly sequence specific manner, only silencing genes to which they are perfectly complementary. This specificity is further supported by the apparent absence of a transitive RNAi pathway in humans.

When considering these results, one should keep in mind that the experiments involved a limited number of siRNAs and cell types. It will be necessary to extend these findings to other cell types and siRNAs--especially those that target endogenous genes. Such experiments are especially critical if siRNAs are to be used as therapeutic agents in the future.

References

1. Elbashir SM, Harborth J, Lendeckel W, Yalcin A, Weber K, and Tuschl T (2001). Duplexes of 21-nucleotide RNAs mediate RNA interference in mammalian cell culture. Nature 411:494-498.
2. Elbashir SM, Martinez J, Patkaniowska A, Lendeckel W, and Tuschl T (2001). Functional anatomy of siRNAs for mediating efficient RNAi in Drosophila melanogaster embryo lysate. EMBO 20 (23): 6877-6888.
3. Chi J-T, Chang HY, Wang NN, Chang DS, Dunphy N, and Brown PO (2003). Genomewide view of gene silencing by small interfering RNAs. Proc. Natl. Acad. Sci. USA (early online edition, 2 May 2003).
   

Related Articles