The importance of RT-qPCR primer design for the detection of siRNA-mediated mRNA silencing
© Reid et al; licensee BioMed Central Ltd. 2011
Received: 28 August 2010
Accepted: 25 May 2011
Published: 25 May 2011
The use of RNAi to analyse gene function in vitro is now widely applied in biological research. However, several difficulties are associated with its use in vivo, mainly relating to inefficient delivery and non-specific effects of short RNA duplexes in animal models. The latter can lead to false positive results when real-time RT-qPCR alone is used to measure target mRNA knockdown.
We observed that detection of an apparent siRNA-mediated knockdown in vivo was dependent on the primers used for real-time RT-qPCR measurement of the target mRNA. Two siRNAs specific for RRM1 with equivalent activity in vitro were administered to A549 xenografts via intratumoural injection. In each case, apparent knockdown of RRM1 mRNA was observed only when the primer pair used in RT-qPCR flanked the siRNA cleavage site. This false-positive result was found to result from co-purified siRNA interfering with both reverse transcription and qPCR.
Our data suggest that using primers flanking the siRNA-mediated cleavage site in RT-qPCR-based measurements of mRNA knockdown in vivo can lead to false positive results. This is particularly relevant where high concentrations of siRNA are introduced, particularly via intratumoural injection, as the siRNA may be co-purified with the RNA and interfere with downstream enzymatic steps. Based on these results, using primers flanking the siRNA target site should be avoided when measuring knockdown of target mRNA by real-time RT-qPCR.
The use of RNAi to inhibit gene expression has revolutionised medical research and has great therapeutic potential. However, inefficient siRNA delivery and off-target effects hamper translation from in vitro experiments to in vivo research and clinic application. Many approaches to improve delivery are under investigation, such as the use of localised delivery by direct injection and topical application, and intravenous administration for systemic delivery [1–3]. Despite the growing use of RNAi in vivo, very few studies include data to confirm that the observed effects of the siRNA are due to an RNAi-mediated mRNA cleavage mechanism rather than non-specific events.
The importance of confirming that mRNA reduction following siRNA administration has occurred via RNAi-mediated events is highlighted by recent studies reporting the contribution of the innate immune system to apparent in vivo knockdown of target mRNAs. The double-stranded nature of siRNA imparts the ability to trigger an innate immune response via the activation of Toll-like receptors (TLR 3, 7 and 8) and binding to proteins such as retinoic acid inducible gene 1 (RIG-1) [4, 5]. These interactions may cause a down-regulation of gene expression that can be falsely attributed to a sequence-specific RNAi-meditated event. Together this suggests that many of the reports of in vivo efficacy of siRNAs can be explained by a general down-regulation of transcription that is stimulated by the double-stranded RNA structure of siRNA without involving RNAi, especially in the absence of corroborating evidence [6, 7].
More recently Holmes et al. found that the 3'fragment produced following siRNA-mediated cleavage of certain target mRNAs can persist and that this can compromise RT-qPCR-mediated detection of knockdown , similar to the findings of others [9, 10] suggesting that incomplete degradation of mRNA cleavage fragments can result in inaccurate determination of knockdown by RT-qPCR. They suggest the use of primers flanking the cleavage site as a means to avoid this problem. Here we show that this approach can lead to artefactual results when siRNAs are used in certain in vivo settings, as siRNAs co-purified with total RNA can interfere with downstream analysis, in some cases leading to false positive results.
Materials and methods
The A549 (human non-small cell lung cancer) and Hepa 1-6 (mouse hepatoma) cell lines used in this study were obtained from ATCC and were grown in RPMI medium supplemented with 10% heat-inactivated fetal bovine serum (FBS) (both from Invitrogen Corporation, Carlsbad, CA), at 37°C in humidified air with 5% CO2.
siRNAs and transfection
Sequences of siRNAs, RNA oligos and RT-qPCR primers used in the study
ApoB1 mm control
Single-strand RNA oligos
5'-RLM-RACE was performed using the GeneRacer kit (Invitrogen) with the manufacturer's instructions modified as previously described . The first round 5'RACE reaction product (1 μl) was used as a template for the MBRACE reaction  using the FastStart TaqMan® Probe Master (Roche) and primers and probes at the following concentrations: 180 nM MB-R, 3.6 μM MB-F and 250 nM molecular beacon probe (all specific for the target gene cleavage site). Reactions were run on a LightCycler 480 as described .
A549 or Hepa 1-6 cells grown in vitro were detached from flasks with trypsin, and the enzymatic reaction was stopped by the addition of fresh culture medium containing FBS. After two washes with PBS, the cells were resuspended in PBS at a final density of 8 × 107 cells/ml (A549 cells) or 5 × 106 cells/ml (Hepa1-6 cells). Using a 26-gauge needle, groups of five CD1 nude mice were injected subcutaneously on the flank with 100 μl of the cell suspension. When tumours reached 50-100 mm3 in size, they were twice injected (24 h apart) with 25 μg siRNA in 50 μl saline; tumours were excised as described (12). The effect of intratumoural injection on RRM1 or ApoB mRNA levels was assessed by real-time RT-qPCR 24 h after the second injection.
Results and Discussion
RRM1 reduction measured by RT-qPCR following intratumoural siRNA administration is primer-pair specific
In a previous screen we identified RRM1 as a potential target for siRNA-based cancer therapy . Three different siRNAs (RRM1-1, -2 and -3) induced significant knockdown of RRM1 mRNA and protein in A549 cells in vitro, leading to growth inhibition and the induction of apoptosis . Furthermore, transfection of A549 cells with RRM1-2 prior to implantation into nude mice markedly inhibited tumour growth . Having observed growth inhibitory effects both in vitro and in vivo following RRM1 knockdown in tumour cells, we then assessed the ability of siRNAs RRM1-2 and RRM1-3 to silence RRM1 mRNA via intratumoural injection of siRNAs into pre-existing subcutaneous xenografts. These siRNAs were shown to be equally potent with IC50 values in vitro of approximately 20 pM [11, 14]. Tumours were twice injected (24 h apart) with 25 μg of RRM1-2 or RRM1-3 siRNA in 50 μl normal saline and excised 24 h after the second injection. RNA was isolated and RRM1 expression analysed by RT-qPCR using two different primer pairs (RRM1-B and RRM1-C in Figure 1).
The presence of siRNA compromises downstream reactions
The apparent knockdown of RRM1/Rrm1 in tumours following intratumoural injection of siRNA was found to relate to the primer pair used in real-time RT-qPCR, in contrast to the similar levels of knockdown measured in vitro after transfection with three different siRNAs, irrespective of the primer pair used in RT-qPCR. This suggested that the residual siRNA was interfering with subsequent steps in the analysis, perhaps explained by the concentration of siRNA in the RNA isolated in each system. The method of RNA isolation used in this study involves binding the RNA to a size selection column, which should exclude small RNAs less than 90 bp. However, miRNAs have been isolated using this procedure  suggesting that the columns only reduce, but do not exclude, small RNAs.
In the in vitro transfection, an siRNA concentration of 10 nM is the equivalent to ~80 ng per well in a 24-well plate. If one assumes cells take up half of the siRNA used in the transfection (and minimal degradation occurs during the 24 h transfection period), the siRNA component of the RNA isolated from the cells (around 10 μg) is less than 1% of the total (80 ng siRNA in ~10 μg of cell-derived total RNA). In contrast, 25 μg siRNA injected twice into a tumour with a volume of 50 mm3 and yielding 50 μg total RNA is likely to be a much greater proportion of the isolated RNA and has the potential to interfere with downstream applications. Further adding to the potential for co-purification is the use of Stealth-modified siRNA duplexes, which have chemical modifications imparting resistance to nucleases and stability in serum.
We also assessed whether siRNA was able to interfere with the reverse transcription step of real-time RT-qPCR. We again used RNA isolated from the tumour of a PBS-treated mouse as template, but here we added increasing concentrations of siRNA or single-stranded RNA to the RNA template in the reverse transcription reaction. RRM1-3 or control siRNA, as well as the single-stranded sense or antisense strands of the RRM1-3 siRNA were added at the concentrations indicated. As seen in Figure 5B, introducing either the RRM1-3 siRNA duplex or antisense strand led to reduced detection of RRM1 message when RRM1-3 flanking primers (RRM1-C) were used. In contrast, introducing RRM1-2 siRNA or RRM1-3 sense strand had no effect on measurements of RRM1 levels. When the primer pair flanking the RRM1-1 site (RRM1-A) was used, no interference was observed (data not shown). Lastly, we added RRM1-2 or RRM1-3 siRNA (25 μg) to tumours prior to RNA extraction, and evaluated mRNA measurements with different primer pairs (Figure 5C). Only when a flanking primer pair was used with template containing the corresponding siRNA was an apparent knockdown detected; no effect was seen with the RRM1-A primer pair, which is located upstream of both RRM1-2 and RRM1-3 sites.
Following intratumoural injection of RRM1-specific siRNAs, the apparent reduction of RRM1 transcript levels was found to be a function of the primer pair used. Subsequent in vitro investigations suggested that this most likely resulted from interference with reverse transcription, and to a lesser extent real-time qPCR, caused by siRNA co-purified in the RNA isolation. These data suggest that primers flanking the siRNA target site should be avoided in studies of siRNA in vivo, especially when large amounts of siRNA are used.
We thank Damian White for help with qPCR, and Marika Eszes and Sheryl Feng for expert assistance with animal work.
- Behlke MA: Progress towards in vivo use of siRNAs. Mol Ther. 2006, 13: 644-670. 10.1016/j.ymthe.2006.01.001.PubMedView ArticleGoogle Scholar
- Dykxhoorn DM, Lieberman J: Running interference: prospects and obstacles to using small interfering RNAs as small molecule drugs. Annu Rev Biomed Eng. 2006, 8: 377-402. 10.1146/annurev.bioeng.8.061505.095848.PubMedView ArticleGoogle Scholar
- Whitehead KA, Langer R, Anderson DG: Knocking down barriers: advances in siRNA delivery. Nat Rev Drug Discov. 2009, 8: 129-138. 10.1038/nrd2742.PubMedView ArticleGoogle Scholar
- Robbins M, Judge A, MacLachlan I: siRNA and innate immunity. Oligonucleotides. 2009, 19: 89-102. 10.1089/oli.2009.0180.PubMedView ArticleGoogle Scholar
- Rossi JJ: Innate immunity confounds the clinical efficacy of small interfering RNAs (siRNAs). Gene Ther. 2009, 16: 579-580. 10.1038/gt.2009.26.PubMedView ArticleGoogle Scholar
- Kleinman ME, Yamada K, Takeda A, Chandrasekaran V, Nozaki M, Baffi JZ, Albuquerque RJ, Yamasaki S, Itaya M, Pan Y: Sequence- and target-independent angiogenesis suppression by siRNA via TLR3. Nature. 2008, 452: 591-597. 10.1038/nature06765.PubMedPubMed CentralView ArticleGoogle Scholar
- Robbins M, Judge A, Ambegia E, Choi C, Yaworski E, Palmer L, McClintock K, MacLachlan I: Misinterpreting the therapeutic effects of small interfering RNA caused by immune stimulation. Hum Gene Ther. 2008, 19: 991-999. 10.1089/hum.2008.131.PubMedView ArticleGoogle Scholar
- Holmes K, Williams CM, Chapman EA, Cross MJ: Detection of siRNA induced mRNA silencing by RT-qPCR: considerations for experimental design. BMC Res Notes. 2010, 3: 53-10.1186/1756-0500-3-53.PubMedPubMed CentralView ArticleGoogle Scholar
- Hahn P, Schmidt C, Weber M, Kang J, Bielke W: RNA interference: PCR strategies for the quantification of stable degradation-fragments derived from siRNA-targeted mRNAs. Biomol Eng. 2004, 21: 113-117. 10.1016/j.bioeng.2004.09.001.PubMedView ArticleGoogle Scholar
- Shepard AR, Jacobson N, Clark AF: Importance of quantitative PCR primer location for short interfering RNA efficacy determination. Anal Biochem. 2005, 344: 287-288. 10.1016/j.ab.2005.06.005.PubMedView ArticleGoogle Scholar
- Reid G, Wallant NC, Patel R, Antonic A, Saxon-Aliifaalogo F, Cao H, Webster G, Watson JD: Potent subunit-specific effects on cell growth and drug sensitivity from optimised siRNA-mediated silencing of ribonucleotide reductase. J RNAi Gene Silencing. 2009, 5: 321-330.PubMedPubMed CentralGoogle Scholar
- Lasham A, Herbert M, Coppieters't Wallant N, Patel R, Feng S, Eszes M, Cao H, Reid G: A rapid and sensitive method to detect siRNA-mediated mRNA cleavage in vivo using 5' RACE and a molecular beacon probe. Nucleic Acids Res. 2010, 38: e19-10.1093/nar/gkp1076.PubMedPubMed CentralView ArticleGoogle Scholar
- Livak KJ, Schmittgen TD: Analysis of relative gene expression data using real-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods. 2001, 25: 402-408. 10.1006/meth.2001.1262.PubMedView ArticleGoogle Scholar
- Patel R, T'Wallant NC, Herbert MH, White D, Murison JG, Reid G: The potency of siRNA-mediated growth inhibition following silencing of essential genes is dependent on siRNA design and varies with target sequence. Oligonucleotides. 2009, 19: 317-328. 10.1089/oli.2009.0207.PubMedView ArticleGoogle Scholar
- Patnaik SK, Kannisto E, Yendamuri S: Overexpression of microRNA miR-30a or miR-191 in A549 lung cancer or BEAS-2B normal lung cell lines does not alter phenotype. PLoS One. 2010, 5: e9219-10.1371/journal.pone.0009219.PubMedPubMed CentralView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.