- Open Access
In vivo quantification of formulated and chemically modified small interfering RNA by heating-in-Triton quantitative reverse transcription polymerase chain reaction (HIT qRT-PCR)
© Landesman et al; licensee BioMed Central Ltd. 2010
- Received: 14 May 2010
- Accepted: 23 August 2010
- Published: 23 August 2010
While increasing numbers of small interfering RNA (siRNA) therapeutics enter into clinical trials, the quantification of siRNA from clinical samples for pharmacokinetic studies remains a challenge. This challenge is even more acute for the quantification of chemically modified and formulated siRNAs such as those typically required for systemic delivery.
Here, we describe a novel method, heating-in-Triton quantitative reverse transcription PCR (HIT qRT-PCR) that improves upon the stem-loop RT-PCR technique for the detection of formulated and chemically modified siRNAs from plasma and tissue. The broad dynamic range of this assay spans five orders of magnitude and can detect as little as 70 pg duplex in 1 g of liver or in 1 ml of plasma. We have used this assay to quantify intravenously administrated siRNA in rodents and have reliably correlated target reduction with tissue drug concentrations. We were able to detect siRNA in rat liver for at least 10 days post injection and determined that for a modified factor VII (FVII) siRNA, on average, approximately 500 siRNA molecules per cell are required to achieve a 50% target reduction.
HIT qRT-PCR is a novel approach that simplifies the in vivo quantification of siRNA and provides a highly sensitive and reproducible tool to measure the silencing efficiency of chemically modified and formulated siRNAs.
- Antisense Strand
- Amplification Curve
- siRNA Duplex
- Broad Dynamic Range
- Total Incubation Time
Several small interfering RNA (siRNA)-based therapeutics are currently in various phases of preclinical and clinical development [1–3]. There is an unmet need to develop sensitive methods to detect and quantify siRNAs in cells and tissues. Therapeutic siRNAs for systemic delivery are typically chemically modified and formulated. Chemical modifications of the sugar-phosphate backbone stabilise siRNA duplexes by enhancing their nuclease resistance and increasing their specificity by reducing off-target effects . Of all delivery systems described, the formulation of siRNAs within cationic lipid nanoparticles (LNPs) is the most validated method for delivery to liver and possibly to other organs . While chemical modifications contribute to increased siRNA stabilisation and specificity, they can significantly increase duplex melting temperature. Moreover, although formulations enhance siRNA tissue delivery and cellular uptake, they can inhibit siRNA release and detection. Currently used siRNA quantification methods include visualisation of radiolabelled siRNAs, hybridisation-based assays with labelled probes using enzyme-linked immunosorbent assays (ELISAs; specifically, hybridisation ELISA), methods based on high performance liquid chromatography (HPLC), capillary gel electrophoresis (CGE) and liquid chromatography mass spectrometry (LC-MS) alone or in combination with hybridisation-based assays (reviewed by Tremblay and Oldfield ). The most sensitive of these bioanalytical methods is the hybridisation ELISA with lower limit of quantification (LLOQ) of approximately 1 ng/ml of a 20-mer oligonucleotide in plasma and approximately 1 ng/g in liver . However, these assays are laborious and have limited sensitivity. A more sensitive assay that is simple and fast to perform and that allows the quantification of therapeutic siRNAs from tissue and plasma is greatly needed.
Recently, Chen et al. developed a stem-loop quantitative reverse transcription PCR (qRT-PCR) method for the quantification of microRNAs (miRs) from tissue culture cells . This assay is particularly appealing because of its high sensitivity, selectivity, and broad dynamic range for the detection of short RNA single strands characteristic of mature miRs. However, when applied to the detection of siRNA duplexes in liver, the stem-loop RT-PCR assay demonstrated a poor dynamic range and suboptimal amplification, suggesting inefficiencies in the RT and PCR steps [9, 10]. In addition, amplification efficiency was lower for chemically modified versus unmodified siRNAs . Such differences may result from the double-stranded nature of siRNAs, their chemical modifications and formulations.
Here, we describe the heating-in-Triton (HIT) method that, when used in combination with the stem-loop qRT-PCR technique, allows robust siRNA quantification from cell lines, tissues and plasma. Using this assay we measured unmodified and chemically modified siRNAs, either as unformulated (naked) siRNAs, or as formulated duplexes within LNPs [11, 12]. Importantly, this assay is strand specific and thus requires complete denaturation of the sense and antisense strands. The assay is highly specific and sensitive, with a broad dynamic range of approximately five orders of magnitude that allows the detection of as little as 70 pg siRNA in 1 g of tissue or 1 ml of plasma. With simple and rapid sample processing, high reproducibility and applicability independent of the chemical modification or delivery strategy employed, this assay is robust and easily adapted for high throughput bioanalytical quantification of siRNA in vivo.
siRNA quantification requires duplex liberation from the formulation followed by strand separation
Small interfering RNA (siRNA) duplexes
To test the effect of strand separation on siRNA detection by qRT-PCR, two siRNA duplexes and their matching antisense strands were heated for 10 minutes at 95°C and directly added into the RT reaction. Under these conditions, similar amplification curves were obtained for duplexes and single strands, particularly at concentrations equal or below 1.7 nmol siRNA per RT reaction (3.35 fmol per PCR reaction, see the -2.5 log siRNA point in Figure 1b), suggesting that duplex siRNAs can be accurately measured when the RT is performed on melted duplexes (Figure 1b).
In vivo delivery typically requires siRNAs formulation in LNPs prior to intravenous injection [11, 13]. LNPs impose an additional barrier to quantification as the siRNA must be efficiently liberated from the particles before quantification. To test our ability to quantify LNP-formulated siRNAs, serial dilutions of the modified siRNA encapsulated in an LNP containing the novel lipid KC2  (KC2 mod Dup), were compared to the modified unformulated siRNA (mod Dup) and the modified antisense strand (mod AS) (Figure 1c). As we had seen previously, unformulated duplex and single strands, heated at 95°C, showed similar amplification curves over at least four orders of magnitude, while formulated duplexes were amplified less efficiently. Therefore, we had to develop a method to disrupt LNP to liberate siRNAs for accurate quantification.
To release the siRNA from the LNP-KC2 formulation, we tested various detergents (data not shown) and ultimately selected 0.25% Triton X-100 for addition to the formulated siRNA prior to heating at 95°C. Under these conditions, nearly identical amplification curves were obtained for the modified antisense strand, modified duplex and LNP-KC2-formulated modified duplex (Figure 1d). A linear regression fit to the three amplification curves over five orders of magnitude showed an average slope of 3.2, which indicated nearly 100% efficient amplification under these modified conditions (Figure 1d).
High temperature is required to avoid reannealing
Quantification of modified and formulated siRNA from plasma and liver tissue
Quantification of siRNA and target expression in rat liver
FVII transcript levels measured in liver by qRT-PCR assay decreased by as much as 80% at 1 day post siRNA administration (Figure 7c) and were tightly correlated with siRNA levels across the three doses and at all time points tested (see Figure 7d; linear regression significant at P < 0.01). Independent of the dose delivered, only 1.6 ± 0.8 ng of siRNA per 1 g of liver was needed to achieve the ED50 of FVII gene silencing (Figure 7d). Assuming 1.5 × 108 cells/g liver tissue , the amount of siRNA at the ED50 is equivalent to approximately 500 siRNA molecules/liver cell. The linear regression also indicates that to achieve 80% target depletion, approximately 15 ± 8 ng siRNA/g liver is required. Such concentrations were attained only on day 1 in the rats that were injected with 0.125 and 0.25 mg/kg of AD1661 (Figure 7c).
Prediction of specificity: recognition of siRNA metabolites
Set of 5' and 3' truncations of the modified antisense A4724.
The HIT siRNA quantification method extends the work of Chen et al., who developed the original stem-loop qRT-PCR method for the quantification of miRNAs . They demonstrated high assay sensitivity and showed its broad utility by measuring the levels of 10 different miRNAs from tissue culture cells, using various sample preparation techniques. Specifically, they observed that quantification from a PBS-boiled tissue culture suspension was as efficient as that from purified RNA, thus eliminating the RNA purification step in their protocol. While the direct boiling/cooling method described by Chen was simple, it was clear that additional steps were needed to make this protocol suitable for the quantification of chemically modified and formulated siRNAs from animal tissue. Importantly, efficient strand separation turned out to be critical for accurate quantification of siRNAs by qRT-PCR. Strand reannealing occurs rapidly in cooled siRNA lysates. As a result, the direct application of heat-denatured siRNAs into the RT reaction was required when measuring high concentrations of high Tm duplexes such as AD1661. In addition, effective siRNA quantification required complete release of the siRNA from the liposomal formulation, a challenge not relevant to endogenous miRNA detection assays. To accomplish this, 0.25% Triton at 95°C was added, as heating in PBS alone was not sufficient to fully liberate the duplex.
The rate of duplex reannealing is directly correlated with duplex concentration and inversely correlated with increased temperature. The highest siRNA concentration in an RT reaction used in our experiments was 17 nmol, from which 33.5 fmol were analysed in one qRT-PCR reaction. At that concentration, reannealing of duplexes occurred rapidly and amplification was not fully efficient, as evidenced by the non-linearity of the curves at this input dose (see the -1.5 points in Figures 1d and 3a). Based on this observation, the upper limit of quantification for our standard curves was defined as 3.4 fmol per qRT-PCR reaction. This translated to approximately 0.7 μg duplex/ml plasma and 0.7 μg duplex/g of liver tissue (Figure 4). The average lowest limit of quantification was 0.34 attmol of siRNA, translated to approximately 70 pg duplex/ml plasma and 70 pg duplex/g of liver tissue, yielding signals that are 4-fold to 10-fold above background signal (a difference of 2-3 Av Ct units).
Having developed an assay for the quantification of formulated and chemically modified siRNA in liver tissue or plasma, the next challenge was to use this assay for the quantification of such siRNAs delivered to rats by intravenous dosing. The goal was to quantify the siRNA antisense strand from siRNA duplex that targets FVII mRNA in rat liver. The antisense strand was selected as it represents the functional targeting strand, but the assay is equally amenable to sense strand detection (Figure 6a). In this in vivo experiment, 0.0625, 0.125 and 0.25 mg/kg of LNP-KC2-formulated AD1661 were delivered as single doses, and samples were quantified for both target mRNA and siRNA levels at 1, 3, 7, 10, or 20 days post dosing. A linear correlation between the amounts of siRNA delivered and target knockdown was observed in the rat livers (Figure 7). This correlation indicated that independent of the dose administered, approximately 1.6 ± 0.8 ng/g of modified AD1661 siRNA resulted in an approximately 50% reduction of FVII mRNA (ED50 = approximately 1.6 ± 0.8 ng/g). This translates to an average of 500 siRNA molecules/cell in rat liver.
To confirm that extraction of siRNA using the HIT qRT-PCR protocol is efficient, we repeated the assay on a subset of the above rat liver tissue samples using a phenol chloroform extraction protocol. The Ct values measured from these two assay protocols were almost identical suggesting that the HIT qRT-PCR assay protocol efficiently liberates siRNAs for easy detection (data not shown). This experiment not only confirmed the quantification results by another method, but also assured us that our extraction method results in efficient release of protein bound siRNA, such as might be bound to the Argonaute 2 (Ago2) protein.
By analysing sets of truncated antisense molecules we aimed to predict the type of siRNA metabolites that one may detect when using the HIT qRT-PCR method. As expected, the assay is more sensitive to 3' end truncations than to 5' end truncations (Figure 8). We observed that siRNA metabolites that include strand truncations of up to five nucleotides at their 5' end and up to two nucleotides on their 3' end are detected by the HIT qRT-PCR method with reasonable efficiency. Future assay development, including modifications of primer length, will further improve HIT qRT-PCR's ability to specifically measure full-length and siRNA metabolites.
In summary, we developed a reliable, reproducible and highly sensitive method of quantitating siRNAs delivered in vivo. As more siRNA therapeutic programmes are initiated and move into preclinical and clinical stages, accurate quantification of siRNA levels across various tissues will become crucially important. This method allows the correlation of pharmacodynamic with precise pharmacokinetic measures, using a sensitive and easy to perform assay that will contribute to the successful development of siRNA therapeutics.
Oligonucleotides and TaqMan probes for the stem-loop real time PCR assays.
(6-FAM) CTGGATACGACAAGGAT (MGB)
(6-FAM) CTGGATACGACAGGGAA (MGB)
(6-FAM) CTGGATACGACAACTTA (MGB)
(6-FAM) CTGGATACGACAAGTAA (MGB)
Quantification of siRNA in liver tissue or plasma in 0.25% Triton: the HIT qRT-PCR protocol (Figure 9)
Following plasma collection and necropsy, plasma and liver samples were kept frozen and liver tissue ground for 2 minutes (min) at 250 strokes/min using a Geno Grinder (SPEXP SamplePrep, NJ, USA) to achieve a light homogeneous powder.
Preparation of standard curves
Duplex, antisense and formulated siRNAs (20 μM) were used for the preparation of serial 10-fold dilutions into 95°C boiled tissue (100 mg/ml) or plasma (1:10 diluted) in 0.25% Triton X-100 in PBS, as detailed below.
For tissue, 500 μl of 0.25% Triton X-100 at 95°C was added directly into 50 mg of naïve (PBS-treated animal) frozen powdered tissue. The lysate was vortexed and put back into the 95°C hot block. Then, 12.5 μl of 20 μM siRNA was added into the hot lysate, which was then vortexed again and put back into the hot block for a total incubation time of 10 min. This procedure was performed once for formulated duplex, naked duplex and for antisense or sense strands. Additional tubes with naïve tissue lysates in 0.25% Triton X-100 were heated at 95°C. These were used for the preparation of the standard curve points. Following 10 min at 95°C, all lysates were cooled on ice for 10 min.
For plasma, plasma from naïve animals (PBS treated) was diluted 1:10 in 0.25% Triton X-100, and 500 μl taken into an Eppendorf tube and heated in a 95°C hot block. Then, 12.5 μl of 20 μM siRNA was added into the hot lysate, which was then vortexed again and put back into the hot block for a total incubation time of 10 min. This procedure was performed once for formulated duplex, naked duplex and for antisense or sense strands. Additional tubes with naïve plasma lysates in 0.25% Triton X-100 were heated at 95°C. These were used for the preparation of the standard curve points. Following 10 min at 95°C, all lysates were cooled on ice for 10 min.
Cooled tissue and plasma lysate tubes were centrifuged at 20,000 g for 20 min at 4°C and supernatants placed into clean Eppendorf tubes and kept on ice.
The naïve and siRNA-spiked (500 nmol), heated and cleared tissue or plasma lysates were used to prepare 10-fold serial dilution points (50 nmol to 50 fmol) for the standard curves.
Preparation of liver and plasma samples for the quantification of siRNA
For tissue, 500 μl of 0.25% Triton X-100 at 95°C was added directly into each 50 mg frozen powdered tissue sample. Lysates were vortexed and put back into the 95°C hot block for a total incubation time of 10 min. Lysates were vortexed twice more during this incubation. Following 10 min at 95°C, all lysates were cooled on ice for 10 min.
For plasma, plasma samples were diluted 1:10 in 0.25% Triton X-100. Then, 500 μl from each diluted plasma sample was heated at 95°C for a total incubation time of 10 min and vortexed twice during this incubation. Following 10 min at 95°C, all lysates were cooled on ice for 10 min.
All tissue and plasma lysates were centrifuged at 20,000 g for 20 min at 4°C and supernatants taken into clean Eppendorf tubes and kept on ice.
Reverse transcription reactions
Reverse transcription reactions were performed using a TaqMan MicroRNA Reverse Transcription kit 200 (Applied Biosystems of Life Technologies, cat # 4366596).
We recommend using two adjacent PCR machines for the procedure. One PCR machine for heating the 'boiling plate' and the second for the 'RT plate', as detailed below.
A total of 50 μl from each standard curve point (single strand, duplex and formulated duplex), one naïve (for background) and sample lysates were aliquoted into one PCR plate ('boiling plate') and heated at 95°C for 10 min on one PCR machine. Then, 10 μl of RT reaction mix (100 mmol deoxyribonucleotide triphosphates, 250 nmol stem and loop oligonucleotides, 20 U/μl RNase inhibitor, 1 × RT buffer, 50 U/μl MultiScribe Reverse Transcriptase) was aliquoted into each well of the 'RT plate', which was placed on the second adjacent PCR machine and kept at 4°C. Following 10 min of heating, the cover from the 'boiling plate' was removed and, while the plate was kept on the heating block, 5 μl from each hot sample was directly added into the RT reaction mix (at 4°C) on the 'RT plate' and the program was switched onto the RT program (30 min, 16°C, 30 min, 42°C, 5 min, 85°C).
A total of 2 μl of cDNA from the previous step was added into the PCR amplification reaction mix (0.2 μM TaqMan probe, 1.5 μM forward primer, 0.7 μM reverse primer, TaqMan 2 × Universal PCR Master Mix, No AmpErase UNG; Applied Biosystems of Life Technologies, cat # 4366596). The ABI 7900HT Sequence Detection System and the 'Standard Curve' application SDS 2.3 (Applied Biosystems of Life Technologies) were used to run the PCR reaction.
Quantification of Gene knockdown in rat liver
An RNeasy mini kit (Qiagen Inc. Valencia, CA, USA, cat # 74106) was used to purify total RNA from frozen powdered rat liver. A High Capacity cDNA Reverse Transcription Kit (Applied Biosystems of Life Technologies, cat # 4368814) was used for RT. TaqMan Assays Rat FVII (cat # Rn00596104_m1) and TaqMan Assay Rat GAPDH kit (cat # 4352338E) together with TaqMan Universal PCR Master Mix No AmpErase UNG (cat # 4324020, all from Applied Biosystems of Life Technologies) were used for the amplification of rat FVII and GAPDH transcripts.
We thank Klaus Charisse for stimulating discussions. We thank Martin Maier, David Bumcrot, Tatiana Novobrantseva, Stuart Milstein, Amy White, Jamie Wong, Hila Epstein-Barash and the RLD group for helpful comments and discussions. We thank Ken Kulmatycki for participating in planning the in vivo experiment. We thank Lauren Lesser for graphic support.
- de Fougerolles A, Vornlocher HP, Maraganore J, Lieberman J: Interfering with disease: a progress report on siRNA-based therapeutics. Nat Rev Drug Discov. 2007, 6: 443-453. 10.1038/nrd2310.View ArticlePubMedGoogle Scholar
- de Fougerolles AR: Delivery vehicles for small interfering RNA in vivo. Hum Gene Ther. 2008, 19: 125-132. 10.1089/hum.2008.928.View ArticlePubMedGoogle Scholar
- Novobrantseva TI, Akinc A, Borodovsky A, de Fougerolles A: Delivering silence: advancements in developing siRNA therapeutics. Curr Opin Drug Discov Devel. 2008, 11: 217-224.PubMedGoogle Scholar
- Jackson AL, Linsley PS: Recognizing and avoiding siRNA off-target effects for target identification and therapeutic application. Nat Rev Drug Discov. 2010, 9: 57-67. 10.1038/nrd3010.View ArticlePubMedGoogle Scholar
- Semple SC, Akinc A, Chen J, Sandhu AP, Mui BL, Cho CK, Sah DW, Stebbing D, Crosley EJ, Yaworski E, Hafez IM, Dorkin JR, Qin J, Lam K, Rajeev KG, Wong KF, Jeffs LB, Nechev L, Eisenhardt ML, Jayaraman M, Kazem M, Maier MA, Srinivasulu M, Weinstein MJ, Chen Q, Alvarez R, Barros SA, De S, Klimuk SK, Borland T, Kosovrasti V, Cantley WL, Tam YK, Manoharan M, Ciufolini MA, Tracy MA, de Fougerolles A, MacLachlan I, Cullis PR, Madden TD, Hope MJ: Rational design of cationic lipids for siRNA delivery. Nat Biotechnol. 2010, 28: 172-176. 10.1038/nbt.1602.View ArticlePubMedGoogle Scholar
- Tremblay GA, Oldfield PR: Bioanalysis of siRNA and oligonucleotide therapeutics in biological fluids and tissues. Bioanalysis. 2009, 1: 595-609. 10.4155/bio.09.66.View ArticlePubMedGoogle Scholar
- Geary RS, Yu RZ, Watanabe T, Henry SP, Hardee GE, Chappell A, Matson J, Sasmor H, Cummins L, Levin AA: Pharmacokinetics of a tumor necrosis factor-alpha phosphorothioate 2'-O-(2-methoxyethyl) modified antisense oligonucleotide: comparison across species. Drug Metab Dispos. 2003, 31: 1419-1428. 10.1124/dmd.31.11.1419.View ArticlePubMedGoogle Scholar
- Chen C, Ridzon DA, Broomer AJ, Zhou Z, Lee DH, Nguyen JT, Barbisin M, Xu NL, Mahuvakar VR, Andersen MR, Lao KQ, Livak KJ, Guegler KJ: Real-time quantification of microRNAs by stem-loop RT-PCR. Nucleic Acids Res. 2005, 33: e179-10.1093/nar/gni178.PubMed CentralView ArticlePubMedGoogle Scholar
- Stratford S, Stec S, Jadhav V, Seitzer J, Abrams M, Beverly M: Examination of real-time polymerase chain reaction methods for the detection and quantification of modified siRNA. Anal Biochem. 2008, 379: 96-104. 10.1016/j.ab.2008.05.001.View ArticlePubMedGoogle Scholar
- Cheng A, Li M, Liang Y, Wang Y, Wong L, Chen C, Vlassov AV, Magdaleno S: Stem-loop RT-PCR quantification of siRNAs in vitro and in vivo. Oligonucleotides. 2009, 19: 203-208. 10.1089/oli.2008.0176.View ArticlePubMedGoogle Scholar
- Zimmermann TS, Lee AC, Akinc A, Bramlage B, Bumcrot D, Fedoruk MN, Harborth J, Heyes JA, Jeffs LB, John M, Judge AD, Lam K, McClintock K, Nechev LV, Palmer LR, Racie T, Röhl I, Seiffert S, Shanmugam S, Sood V, Soutschek J, Toudjarska I, Wheat AJ, Yaworski E, Zedalis W, Koteliansky V, Manoharan M, Vornlocher HP, MacLachlan I: RNAi-mediated gene silencing in non-human primates. Nature. 2006, 441: 111-114. 10.1038/nature04688.View ArticlePubMedGoogle Scholar
- Akinc A, Goldberg M, Qin J, Dorkin JR, Gamba-Vitalo C, Maier M, Jayaprakash KN, Jayaraman M, Rajeev KG, Manoharan M, Koteliansky V, Röhl I, Leshchiner ES, Langer R, Anderson DG: Development of lipidoid-siRNA formulations for systemic delivery to the liver. Mol Ther. 2009, 17: 872-879. 10.1038/mt.2009.36.PubMed CentralView ArticlePubMedGoogle Scholar
- Morrissey DV, Lockridge JA, Shaw L, Blanchard K, Jensen K, Breen W, Hartsough K, Machemer L, Radka S, Jadhav V, Vaish N, Zinnen S, Vargeese C, Bowman K, Shaffer CS, Jeffs LB, Judge A, MacLachlan I, Polisky B: Potent and persistent in vivo anti-HBV activity of chemically modified siRNAs. Nat Biotechnol. 2005, 23: 1002-1007. 10.1038/nbt1122.View ArticlePubMedGoogle Scholar
- Love KT, Mahon KP, Levins CG, Whitehead KA, Querbes W, Dorkin JR, Qin J, Cantley W, Qin LL, Racie T, Frank-Kamenetsky M, Yip KN, Alvarez R, Sah DW, de Fougerolles A, Fitzgerald K, Koteliansky V, Akinc A, Langer R, Anderson DG: Lipid-like materials for low-dose, in vivo gene silencing. Proc Natl Acad Sci USA. 2010, 107: 1864-1869. 10.1073/pnas.0910603106.PubMed CentralView ArticlePubMedGoogle Scholar
- Hendriks HF, Brouwer A, Knook DL: Isolation, purification, and characterization of liver cell types. Methods Enzymol. 1990, 190: 49-58. full_text.View ArticlePubMedGoogle 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.