by Laura Lane
The tide has certainly turned for RNA. Finally receiving due recognition for its integral role in regulating gene expression, RNA can at last shed its reputation as merely “a humble carrier of messages and fetcher of building materials,” as described in a recent issue of The Economist1, which published a special report on life science’s latest revelations of RNA’s functions. Such findings are beginning to change “people’s views about how cells regulate themselves, how life becomes more complex, how certain mysterious diseases develop and even how the process of evolution operates.”
siRNA and RNA interference
At the very least, RNA-focused research has turned up an irreplaceable tool for studying gene expression: siRNA, also known as small interfering RNA. The oligonucleotide participates in the natural process of RNA interference (RNAi) that regulates the translation of mRNA. By synthetically producing siRNAs that are targeted toward specific mRNAs, researchers can study the purpose and function of the related gene. With vast numbers of mRNAs to examine, researchers have turned to the convenience and expertise of custom oligonucleotide services.
In theory, producing siRNAs should be quite straightforward. The foundational composition is its primary structure of 20 to 25 nucleotides with a sequence that will hybridize to specific sites on the targeted mRNA. Once the siRNA sequences hybridize to the mRNA, the complex aggregates into endoribonuclease-containing complexes called RNA-induced silencing complexes (RISCs). After ATP-generated unwinding, the siRNA guides the RISC to the complementary mRNA molecules. The RISC destroys the target, thus preventing translation.
Optimal siRNA design
However, this sequence of events demands much more than solely the correct nucleotide sequence. Designing efficient and effective siRNAs requires software that takes into account numerous parameters and very complex computations. Typically, the programs will ask you to input the target’s nucleotide sequence, the GenBank accession number, the number of oligonucleotide candidates that you’d like to see in the results, any design rules or sequence motifs, the oligonucleotide length that you desire, and the minimum and maximum GC content that works for you.
Then, behind the scenes, the program uses its computational algorithms to calculate and churn out the optimal siRNAs. Embedded in the algorithms are various parameters in maximizing the oligonucleotide’s ability to find and bind to the target. That means programs will be guided by the thermodynamics that dictate siRNA-mRNA binding stability, certain sequence motifs, and the potentials for the siRNA oligonucleotide or the target mRNA to form a secondary structure.
The various parameters, or molecular signatures, guide the program in finding “sequences that will have predicted high activities and avoid sequences with predicted low activities,” says Andy Peek, director of bioinformatics at Integrated DNA Technologies, Inc. “The design programs are computational winnowing machines with the job of separating the grain from the chaff.”
The advanced bioinformatics have yielded oligonucleotides with plenty of targeting power. Still, the correlation between the predicted activity of siRNA, and that which is actually observed, varies from a coefficient of 0.6 to 0.8, depending on the algorithmic model. “We certainly aren’t done in making model improvements,” Peek says.
Further refinement will depend on a more intimate understanding of the mechanism of RNAi, he says. “An algorithm is simply a process description or a model of a series of events, so improvement would involve a more precise model for events like transfection, cellular uptake, RISC loading, RISC maturation, target scanning, catalytic turnover and all the process that we know needs to occur to get to our current measuring stick end point.”
Manipulating pyrimidine and purine content is one common method for altering the oligonucleotides’ efficiencies. Generally, GC content of the entire oligonucleotide should fall between 30% and 50%. And these GC pairs should enrich the 3’ end, according to so-called end rules for efficiency, which also recommend AU-rich five prime ends. But, as usual, exceptions come with every biological rule.
“We do find that some oligonucleotides that fall outside the guidelines can still have high levels of activity,” says Kristin Wiederholt, research and development manager in Invitrogen’s gene expression profiling group.
You can also vary the duplex length and maintain a high level of activity. However, employing duplexes longer than 30 base pairs is not recommonded because they will be recognized as viral RNA and induce an interferon response, Wiederholt says.
More comprehensive silencing of specific genes is not the only goal for maximizing target efficiency. Part of the purpose is to minimize binding to unintended targets. siRNAs face a large pool of RNAs, which offer ample opportunity for erroneous binding. Researchers are also finding that siRNAs can even bind to DNA. Controlling unintended binding events is crucial in harnessing the molecules for drug treatment.
“So there do appear to be several forces acting against specificity of small RNAs,” Peek says. To improve specificity, he suggests, choose “the most active siRNA sequence possible and use it at the lowest effective concentration.”
Custom oligo synthesis services
For the most part, researchers can leave it to the supply companies to grapple with the details. The custom services provide user-friendly software with straightforward interfaces. You can also avail yourself of online tutorials that walk you through the ordering process. You may want to try the many online programs offered by public institutions. Choosing the one that works best for your project will require some investigating.
For open source software, you have the opportunity to inspect the underlying code. But if you’re not so technically oriented, consider programs that were validated using the same types of cells and/or model organisms that you’re studying. Keep in mind that “most siRNA design programs have been modeled on data collected predominantly from mammalian cells—human and mouse—and from a relatively small number of cell types,” Peek says. In addition, look for programs that were created as solutions to problems similar to those you’re confronting.
Once you’ve selected which oligonucleotide candidates to purchase, you can add other specifications to your order. You can choose from several levels of purity, each of which are appropriate for different applications. You can also request chemical modifications to single nucleotides within the siRNA strand. Options include the substitution of DNA bases or other modified nucleic acids, the addition of fluorescent tags or spacers, and many other modifications.
The increasing sophistication of siRNA oligonucleotides is allowing researchers to further their ribonomic-driven studies. Defined as the “identification and analysis of linked mRNA subsets by using RNA-associated proteins” in a paper published in a December 2000 issue of Proceedings of the National Academy of Sciences2, the ribonome is no doubt starting to take form.
References:
1“Really New Advances,” The Economist, 383(8533):87-89, June 16, 2007.
2SA Tenenbaum SA, et al., "Identifying mRNA subsets in messenger ribonucleoprotein complexes by using cDNA arrays," Proceedings of the National Academy of Sciences, 97(26):14085-14090, December 19, 2000.
