Jeffrey Perkel has been a scientific writer and editor since 2000. He holds a PhD in Cell and Molecular Biology from the University of Pennsylvania, and did postdoctoral work at the University of Pennsylvania and at Harvard Medical School.
In the 15 years since Fire and Mello introduced the world to RNA interference (RNAi), the technique has become a standard genetics tool. Researchers who want to quickly gauge the impact of knocking down a gene can skip laborious gene knockouts. Just order up a couple of short oligonucleotides, transfect them into cells and go.
Unlike traditional knockouts, which physically delete or destroy the gene sequence itself, RNAi targets its output, the mRNA. The process is triggered by a double-stranded 21-mer RNA, one strand of which is complementary to the target mRNA. The duplex is loaded into the RNA-induced silencing complex (RISC), and then one strand (the “passenger strand”) is evicted to produce the mature complex. It is this form of the RISC that does the actual work of cutting the targeted mRNA.
In some cases, the double-stranded RNA, called a small-interfering RNA (siRNA), is transfected into the cell directly. In other cases, a longer precursor called a “Dicer-ready siRNA” is added instead, which the cell then processes to produce a mature siRNA. Alternatively, researchers can express a short-hairpin RNA (shRNA) inside the cell from either a plasmid or viral vector, which again is processed to make the mature RNAi trigger.
(An alternative strategy involves mRNA cleavage via an RNase H-mediated pathway. Exiqon’s custom, single-stranded GapmeR oligonucleotides, for instance, so-named because they contain two or three locked nucleic acid (LNA) nucleotides at either end of the molecule flanking a DNA center, can drive this process, but other non-DNA flanking nucleotides will work too, including 2’-O-methyl or 2’-O-methoxyethyl bases.)
Whatever the technical particulars, the process couldn’t be simpler, says Phillip Zamore, the Gretchen Stone Cook Professor of Biomedical Sciences at the University of Massachusetts Medical School. “At this point it’s not so terribly different from using restriction enzymes,” Zamore says. “It’s a very straightforward, simple technique, which I think is why it’s so popular.”
Indeed, a number of companies already offer libraries of pre-designed siRNAs for all or at least most human, mouse and/or rat coding sequences. Life Technologies, for instance, offers three such libraries: the Silencer® siRNA, Stealth RNAi™ siRNA, and Silencer® Select siRNA libraries, which differ in potency, chemical modification and design, among other parameters. (Some other vendors of genome-wide siRNA, shRNA or expression libraries are Open Biosystems, now available through Thermo Fisher Scientific; Sigma-Aldrich; Integrated DNA Technologies (IDT); and Qiagen.)
Still, there are occasions when you’ll need to design your own siRNA. Perhaps you want to knock down a gene in an organism that isn’t so widely studied. Maybe you want to target one of the newly discovered long noncoding RNAs (premade siRNA libraries typically target only coding sequences) or a specific gene isoform. Or maybe the commercially available premade siRNAs just aren’t working for you. Whatever the reason, you’re in luck: It’s relatively simple to design a custom siRNA.
Design rules
When designing siRNAs, site selection is obviously key. You want an siRNA that targets your gene of interest (preferably within the coding sequence itself, as opposed to in the 5’ or 3’ untranslated regions) while avoiding off-targets with related sequences. “There are typically thousands of potential siRNAs you could design for a given gene, and some will work and some won’t,” says Mark Behlke, CSO at IDT, a custom oligonucleotide synthesis company.
The available sequence space is actually quite narrow, says Beverly Davidson, the Roy J. Carver Biomedical Research Chair of Internal Medicine at the University of Iowa College of Medicine. Although an siRNA is about 21 bases long and requires perfect complementarity with its target, microRNAs, which silence genes via a different mechanism, require only a seven-nucleotide “seed,” a length that is guaranteed to be present more than once in a eukaryotic genome. Thus, even the most tightly tailored siRNA sequence could potentially, and inadvertently, shut down thousands of genes.
Fortunately, not all heptanucleotide sequences are present at equal frequencies in the genome, and Davidson and her team have developed an algorithm to identify them. She recently codified those rules in an online server, siSPOTR, to help design siRNAs “with a low potential for off-targeting.”
As for the rest of the sequence, avoid secondary structure in both your siRNA design and target (Zamore recommends Mfold). And, if you intend to use your siRNA in vivo, consider modifications to reduce nuclease stability and immunogenicity (for instance, 2’-O-methyl modifications).
“In general we will recommend more extensive modification for use in vivo than in cell culture,” says Behlke.
Behlke recommends trying multiple siRNAs per gene, just to be sure at least one works. That’s the theory behind IDT’s TriFECTa® RNAi kits, each of which includes three 27-nucleotide Dicer substrate molecules (out of 10 the company has pre-designed for each gene).
Yet according to Zamore, there’s really only one thing to remember when it comes to siRNA design: “thermodynamic asymmetry.” Design your siRNA so that one side unwinds more easily than the other, and you can ensure that the correct strand is loaded into the RISC complex.
Zamore’s lab accomplishes that by a trick he calls “fraying”: Basically, make sure that the 5’-most nucleotide of the antisense (or guide) strand (that is, the strand you want to load into RISC) is mismatched, and chances are you’ll have a working molecule with minimal off-targeting, says Zamore. (Zamore receives royalty payments through the University of Massachusetts for his part in the discovery of siRNAs.)
Experimental control
Several reviews cover the art of designing effective siRNAs, including articles from Davidson [1], Behlke [2] and Zamore’s University of Massachusetts colleague, Anastasia Khvorova [3]. Numerous online tools are available, as well, including from IDT, Life Technologies, and Thermo Fisher Scientific.
Ultimately, though, design is only part of the process; if you want to know if your siRNA is good, you’ll need to test it. And for that, you’ll need the proper controls, both positive and negative.
“Probably the biggest problem we see is that people aren’t getting good transfection, and they don’t even know it,” says Behlke, who recommends an HPRT (hypoxanthine phosphoribosyltransferase) positive control for transfection efficiency.
The problem can seem dizzying: With thousands of different kinds of cells, countless potential siRNA sequences and multiple delivery reagents available (including reagents dedicated to small RNAs, such as Life Technologies’ Lipofectamine® RNAiMAX and Sigma-Aldrich’s MISSION® siRNA Transfection Reagent), optimization is a must when it comes to RNAi.
But you need not be intimidated, says Zamore. “Of all the things in science to be put off by, this should rank near the bottom. It’s just painless.”
References
[1] Davidson, BL, McCray Jr., PB, “Current prospects for RNA interference-based therapies,” Nat Rev Genet, 12:329-40, 2011.
[2] Peek, AS, Behlke, MA, “Design of active small interfering RNAs,” Curr Opin Mol Ther, 9:110-8, 2007.
[3] Birmingham, A, et al., “A protocol for designing siRNAs with high functionality and specificity,” Nat Protocols, 2[9]:2068-78, 2007.
Image credit: iStockPhoto.com