Caitlin Smith has a B.A. in biology from Reed College, a Ph.D. in neuroscience from Yale University, and completed postdoctoral work at the Vollum Institute.
RNA interference (RNAi) is simple in concept yet nuanced in implementation. Key to making experiments work is optimization. The right conditions can take your RNAi experiments from puzzling to perfect. Here are some tools and tips to help.
New delivery tools
Delivering inhibitory RNA molecules such as small/short inhibitory RNA (siRNA), small/short hairpin RNA (shRNA), microRNA (miRNA) and noncoding RNA (ncRNA) is getting easier and more efficient with the emergence of new, dedicated molecular tools.
siRNA
For instance, Polyplus Transfection’s INTERFERin®-HTS platform is designed for high-throughput siRNA transfection experiments using automated liquid-handling systems. High-throughput studies often “require the transfection complexes to be stable for several hours,” says Polyplus’ CEO Mark Bloomfield, with “highly stable transfection complexes of siRNA oligo and transfection reagent combined.”
Polyplus also has developed two modified forms of siRNAs for in vivo transfection. STICKYSIRNA™ increases the stability of the complex formed between the siRNA and the delivery reagent, in vivo-jetPEI®. The second modified siRNA, SIRNAPLUS™, requires no delivery reagent because “the ‘delivery vector’ is built into the siRNA molecule itself,” says Bloomfield.
Another innovative siRNA design, from Integrated DNA Technologies (IDT), is the Dicector library of pre-designed siRNA sequences, an addition to the company’s line of Dicer-substrate siRNAs (DsiRNAs). Dicector DsiRNAs, which target all human, mouse and rat genes, use “a new, improved design algorithm, based on support vector machine methods, that was ‘trained’ from many thousands of data points obtained from large screens,” says IDT chief scientific officer Mark Behlke. Designs from the new algorithm will launch in late Fall 2013.
shRNA
Although siRNAs are typically transiently transfected, shRNAs are usually expressed stably using viral or bacterial vectors. Stable expression of an shRNA is convenient, but random integration into the host cells’ genome can cause shRNA expression levels to vary.
To prevent this, Sigma Aldrich developed a kit that lets you insert the shRNA into a specific spot in the genome. The company’s target integration kit uses zinc finger nuclease technology “to deliver an shRNA to the ‘safe harbor’ AAVS1 locus in the human genome,” says Shawn Shafer, functional genomics market segment manager at Sigma Life Science. Pinning down the shRNA to one site means genomic location is no longer a variable in shRNA expression.
Thermo Fisher Scientific is helping researchers find the best expression promoters for shRNAs. “Poor knockdown due to suboptimal promoter performance might be interpreted as a nonfunctional shRNA or poor transduction efficiency, when in fact it is simply that the shRNA is not being expressed at appropriate levels,” says Louise Baskin, senior product manager at Thermo Fisher Scientific.
The new Thermo Scientific SMARTchoice shRNA platform lets users compare the performance of many promoters in one cell type using a GFP-expressing, non-targeting shRNA. The optimal promoter—i.e., that which produces the strongest GFP signal—can then be incorporated into a targeting shRNA construct and controls for RNAi experiments.
Antisense RNA
IDT also offers tools to knock down the expression of nuclear non-coding (nc) RNAs. Sometimes siRNAs don’t work well for ncRNA knockdown. In these cases, says Behlke, gapmer antisense oligos (ASOs) often do. “This might relate to degradative RNAi being more of a cytoplasmic [process] while degradative RNase H antisense may be more of a nuclear process,” he says. “The sole drawback is that there is not a good antisense design tool available today of the same quality that exists for siRNA, so you have to empirically test more sites when you use ASOs.”
Eric Lader, senior director of product development at QIAGEN, has seen growing interest in RNAi tools targeting long ncRNAs (lncRNAs), along with reverse transcriptase (RT)-qPCR assays to analyze the results. “This requires some advanced bioinformatics, as the lncRNA sequence databases are nowhere near as mature as those for protein-coding genes,” he says.
Verifying knockdown
It almost goes without saying that you should begin an RNAi protocol with healthy cells. Bloomfield recommends using cells that have undergone fewer than 20 passages. And although the type of cells will be dictated by your experimental plans, your RNAi delivery tools may be dictated by the cell type. “There is not a single delivery method that is better than others across all cell lines, so researchers must explore their options when getting started with a new cell type,” says Baskin.
In any event, your goal is to balance RNA delivery efficiency with cytotoxicity.
After you have chosen your inhibitory RNA, it’s important to verify knockdown by measuring both mRNA and protein levels and then comparing to controls. The negative control is typically a nontargeting RNA, and the positive-control RNA usually targets a “housekeeping gene” such as GAPDH, PPIB or HPRT, that you can measure by PCR.
But first, it’s best to make sure your PCR technique is up to snuff. “Using well-characterized positive and negative controls and validated RT-[q]PCR assays is essential to quantify knockdown reproducibly,” says Lader. “The researcher also has to be able to perform RT-PCR reasonably well. The difference between an 80% knockdown and a 50% knockdown is only a fraction of a cycle in real-time PCR and can be difficult to measure.”
Keep the timing of expression in mind, too, when verifying knockdown. “Don’t forget to check the half-lives of the proteins and mRNAs of interest and measure gene silencing accordingly, between 24 to 96 hours after transfection, in order to determine the optimal time window for analysis,” says Bloomfield.
Baskin cautions that a common pitfall in RNAi experiments is for researchers to expect that their targets will be knocked down to the same degree as the housekeeping gene used as a positive control. “Not all target genes will be expressed in the same manner, nor will all RNAi reagents have equivalent potency to a validated positive control,” she says.
Efficacy across cell lines differs, too. “There is a growing body of data to indicate that the same RNAi reagent, when used in different cell types, may have very different results in terms of target-gene silencing as well as specificity and degree of off-target [effects],” Baskin says.
In other words, there is no ideal gene silencer—if you switch cell types, you should not assume that a given siRNA or shRNA's performance will be consistent with results in other cells. “Optimization needs to be done for every cell line and organism that RNAi is being conducted in,” says Shafer. “One cannot assume that knockdown levels will be identical or even comparable when changing systems.”
Variations in transfection efficiencies can skew your results, especially between cell lines and between different types of knockdown reagents. In fact, transfection efficiency can vary from one day to the next even in the same cell line. The best solution is to run both positive and negative controls each time you run a gene-knockdown experiment. “It is just not good enough to have established that a protocol worked once,” says Behlke. “You need the controls run in parallel for every experiment, including a positive control with known potency.”
Although the right controls are critical, it also is advisable to double-check your knockdown results with another set of inhibitors. “It is important to see more than one siRNA or shRNA having the same effect against the same transcript,” says Shafer. “Having tight designs and well-annotated gene sequence data is an underappreciated aspect of conducting a good RNAi experiment.”
Image: Transfection of 1 nM siRNA with Polyplus Transfection's INTERFERin® reagent.