Use of Enzymatically Generated and Chemically Synthesized siRNA

RNA Interference (RNAi), the transfection of nucleic acid into cells to knock down the expression of a specific gene of interest, has proven to be a very powerful tool in the exploration of gene function. One of the most popular and easiest methods to perform RNAi is the use of small interfering RNA (siRNA). These molecules are short double-stranded RNA duplex molecules with lengths that vary across species (15-21 nucleotides). They also generally contain 5' phosphates and 3' two-nucleotide overhangs. The most effective siRNA molecules share 100 percent identity with the targeted gene and no significant homology with any other gene. Other chemical properties of siRNA important for optimal function have been and continue to be defined (1). Researchers currently employ two different methods for the production of siRNA: enzymatic generation or chemical synthesis. This article briefly discusses the difference between these two methods as well as the advantages and disadvantages of each. Experimental results in this article also demonstrate that enzymatic generation may cause fewer non-specific effects than chemical synthesis.

Chemical Synthesis of siRNA:

The chemical synthesis of siRNA utilizes time-tested methods of solid-phase support chemistry for generating oligomers of deoxyribonucleotides such as PCR primers. For siRNA, the same core facilities and commercial providers having access to the required equipment synthesize oligomers of ribonucleotides having complementary sequences. The oligos are then mixed allowing them to anneal, lyophilized and shipped. The required oligo sequence derives from a short 19- to 21-nt segment of the desired target gene. Designers use sophisticated computer algorithms to define the optimal siRNA sequence for the targeted gene. These programs have been experimentally verified using training sets of siRNA directed toward a small subset of genes. For optimal efficacy, designers define and use as many as four different segments of the targeted gene sequence for the chemical synthesis of siRNA.

Several manufacturers offer services to chemically synthesize oligo for siRNA with high purity and guaranteed sequences. However, the development of the computer algorithms and the synthesis of multiple RNA oligos significantly increase the price. Although the computer programs rely on experimentation initially, the manufacturer does not necessarily test each designed and synthesized siRNA by direct experimentation leaving the task to individual end-user. However, the researcher can easily order and screen several siRNA duplexes at the same time. Also, optimal siRNA efficacy against the target gene often requires high concentrations of oligos potentially causing non-specific effects. (See below.)

Readily available from several manufacturers
Purity and oligo sequence always consistent
Multiple potential siRNA duplexes easily screened

High cost and long preparation time
Each duplex or pool of duplexes must be tested for efficacy
Used at a relatively high concentration

Enzymatic Generation of siRNA:

The enzymatic generation of siRNA requires a cloned gene-specific fragment of the targeted gene or a clone of the full-length cDNA (2). This plasmid serves as a substrate in an enzymatic process involving multiple steps to generate a population of siRNA molecules. Amplification of the gene or fragment using primers containing flanking sequence and the T7 promoter generates a double-stranded DNA containing the T7 promoter on both sides to allow transcription of both strands. A simple in vitro transcription reaction using T7 RNA polymerase followed by DNAse treatment yields a long piece of double-stranded RNA. Treatment with the E. coli enzyme RNase III or an appropriate source of the Dicer enzyme randomly cleaves the RNA into a population of siRNA molecules. This method does not rely on design parameters or even necessarily the full sequence information of the targeted gene. The use of a population of duplexes is also similar to the endogenous mechanism of RNAi where the Dicer enzyme digests a long duplex between message and antisense message into many siRNA molecules.

A few companies offer inexpensive kits for enzymatically generating siRNA for a gene of interest; however, cloning the gene or gene-specific fragment requires more time and money to complete. Others companies, such as SABiosciences, prepare gene-specific siRNA population, validate them with cell-based assays, and provide them to researchers in equally inexpensive kits. No special equipment (such as an oligonucleotide synthesizer) is required to enzymatically generate siRNA. The use of a population of siRNA duplexes suggests that this method may be less specific than carefully designed and sequence-based oligos; however, the experiment below suggests the opposite might be true.

Low cost and one-day preparation time
Mimics the endogenous process
Can be performed in any laboratory setting and only one population screened
Knowledge of design parameters or complete sequence information not required

Process requires researcher to perform multiple enzymatic and purification steps
Difficult to increase siRNA yield

Comparing the "Off Target Effect" of the Two Methods:

A recent and very eloquent published study (3) attempts to ascertain whether siRNA designed for a specific gene non-specifically affects the expression of any other genes in the genome. The group chemically synthesizes siRNA directed toward the MAPK14 gene and transfect it into cells. They then isolate RNA and used it for a genome-wide microarray study. Interestingly, the siRNA decreases the expression not only of the targeted gene but also a number of other genes with the same kinetics. This result suggests that the effect is mediated by the siRNA directly and not by the down-regulation of MAPK14, which would have occurred with a delayed time course. Examination of the other affected genes reveals sequences nearly identical to the synthesized siRNA with only one or two substitutions. The results of the paper suggest that short (15-21 nt) double-stranded RNA oligos may not be quite unique enough to target only one gene in the entire genome.

However, the paper also shows that the magnitude of off-target non-specific effect decreases with smaller amounts of transfected siRNA. This observation suggests that populations of enzymatically-generated siRNA may have a distinct advantage over chemically synthesized siRNA. Each duplex in the population has a different sequence and probably has a different off-target effect (non-specifically targets a different set of genes) based on their sequence. However, the individual duplexes in the population represent only a small fraction of total siRNA concentration. Their concentration may small enough not to elicit the non-specific effect, or each duplex may have at least a minimal off-target effect. Since each duplex completely matches the target gene, each should have an additive effect on targeted gene. As a result, enzymatically-generated siRNA may have a reduced off-target effect.

Figure1: Reduced "Off-Target" Effect with Enzymatically-Generated siRNA. HEK-293T cells were transfected with either a negative control (GFP) SureSilencing siRNA population (C, at 20 nM), or a MAPK14 SureSilencing siRNA population from SABiosciences (at 5 and 20 nM), or four chemically synthesized MAPK14 siRNA duplexes (at 5 and 20 nM). RNA was harvested from cells 24 h post-transfection and used to analyze the expression level of four different genes (MAPK14, RRAD, CTGF, and GAPD.) by RT-PCR.

The experiment in Figure 1 tests and verifies this hypothesis. In an experiment modeled on work described above, two commercial sources of siRNA directed towards the MAPK14 gene were transfected into cells at two different concentrations, an enzymatically generated population using a gene-specific fragment and four chemically synthesized oligos having sequences derived from the same gene-specific region. The expression of the target gene and two other genes identified in the previous work, relative to control-transfected cells, was determined by semi-quantitative RT-PCR.

The results demonstrate that both siRNA reagents (at 20 nM) decrease the expression of the targeted MAPK14 gene by at least 70 percent. Neither siRNA method affects the expression of GAPD. However, the synthetic siRNA also non-specifically decreases the expression of RRAD and increases the expression of CTGF, even at the lower concentration (5 nM). The enzymatically-generated population does not affect the expression of these genes even at the higher concentration (20 nM). The reproducible results achieved for this single gene support the hypothesis that enzymatically generated siRNA more specifically knocks down the expression of the targeted gene than chemically synthesized siRNA.