Quantitative Analysis of Dicer Substrate Oligonucleotides in Mouse Liver by Ultra-High-Performance Liquid Chromatography–Electrospray Ionization Tandem Mass Spectrometry

Abstract

A method using multi-mode solid-phase extraction and ultra-high-performance liquid chromatography (UHPLC)–electrospray mass spectrometry was developed to quantify Dicer-substrate small interfering RNA (DsiRNA) directed against the hypoxanthine phosphoribosyltransferase 1 (HPRT1) gene transcript in mouse liver tissue. The oligonucleotides were separated into sense and antisense strands using a UHPLC C₁₈ column with mobile phases containing 1,1,1,3,3,3-hexafluoro-2-propanol in both water (mobile phase A) and methanol (mobile phase B) with triethylamine as the ion pairing agent at a column temperature of 65°C. The lower limits of detection for the sense and antisense strands were ∼1 ng/mg. The dynamic ranges for the sense and antisense strands were 5 ng/mg–1,000 ng/mg and 1 ng/mg–1,000 ng/mg, respectively. The lower limits of quantification for the sense and antisense strands were 5 ng/mg and 1 ng/mg, respectively, each with a relative standard deviation <15% over the range of calibration curve with sufficient precision and accuracy. Oligonucleotides were quantified at different time intervals after intravenous administration of living mice with lipid nanoparticle formulated HPRT1 DsiRNA and were detected as early as 15 min after administration, but not detected beyond the 24 h time point.

Introduction

RNA interference (RNAi) is a naturally occurring cellular process that silences specific genes.^1
–3 The ability to trigger RNAi in mammalian cells using double-stranded small interfering RNA (siRNA) has stimulated great interest in its use as a tool in molecular biology and for therapeutic applications. The natural initiation point for the RNAi pathway is the Ribonuclease enzyme Dicer, which processes double-stranded RNA into siRNAs.^4,5 These siRNAs are typically 19bp in length, with two base 3′ overhangs on each end of the molecule. Dicer-Substrate RNA (DsiRNA), which require processing by Dicer and which are 25–30bp in length containing a single two-base overhang on 3′ end of the antisense strand, has been found to be a more potent and longer-lasting inducer of RNAi than siRNA.⁶ The use of DsiRNA may help in the assembly of the RNA induced silencing complex (RISC), the protein complex that mediates the RNAi mechanism.⁷ After processing by Dicer, the resulting siRNAs are delivered to the RISC loading complex, which separates the sense and antisense strands. The sense strand is subsequently discarded and degraded. The antisense strand is retained and becomes a guide sequence within RISC that binds to complementary messenger Ribonucleic Acid (mRNA) sequences present within the cellular cytoplasm. Argonaute 2, a component of RISC, cleaves the bound mRNA, destroying the transcript and resulting in gene silencing. Due to this highly selective ability to target the destruction of specific gene transcripts, RNAi is being used to develop gene-targeting therapies for diseases such as ovarian cancer,⁸ bone cancer,⁹ hypercholesterolemia,¹⁰ liver cirrhosis,¹¹ respiratory syncytial virus¹² hepatitis B virus,¹³ and others.

Preclinical and clinical development of therapeutic oligonucleotides requires the ability to measure oligonucleotides quantitatively in biological matrices in order to implement pharmacokinetic and toxicokinetic studies. A variety of analytical techniques are currently used for this detection and quantification. For example, high-performance liquid chromatography (HPLC) based on ion-pair reversed phase¹⁴ and anion exchange¹⁵ methods has been applied for the detection and quantification of oligonucleotides in biological matrices. HPLC-based methods are robust and easy to use, but the detection and quantification of oligonucleotides is limited due to the low sensitivity of the UV-based detection.^16,17 Higher HPLC sensitivity has been achieved through florescence detection.¹⁸ Capillary electrophoresis (CE) along with ultraviolet (UV) detection has been successfully used to study oligonucleotides in both plasma and tissues,¹⁹ but again, the low sensitivity of UV detection limits its usefulness in Pharmacokinetics (PK) studies. The use of a lazer-induced florescence detection system can increase the sensitivity of CE up to threefold, compared with the UV detection system, enhancing usefulness.²⁰ Alternatively, hybridization methods based on noncompetitive binding provide both high sensitivity and specificity and can be implemented at high throughput with minimum sample clean-up.^21,22 However, hybridization-based assays tend to have narrow dynamic range, typically <2 orders of magnitude,²² when compared with HPLC or CE based methods just mentioned. Finally, a quantitative polymerase chain reaction provides the highest sensitivity of all current analytical methods,²³ and may not require sample purification.²⁴ Unfortunately, it lacks the selectivity required for metabolite quantification.

Liquid chromatography coupled to tandem mass spectrometry as a quantitation method has been widely used in small-molecule drug discovery and development.²⁵ Due to its high sensitivity and ability to precisely measure masses in the oligonucleotide range, we sought to develop a method that uses ultra HPLC coupled with electrospray ionization mass spectroscopy (UHPLC-ESI-MS) to identify and quantify oligonucleotides in biological matrices. The ability to precisely measure the mass of an oligonucleotide by ESI-MS should allow the confirmation of the presence of an oligonucleotide of a specific elemental composition in a biological sample, as well as an assessment of its purity and the nature of any contaminants, modifications, and metabolites. However, the use of MS for the detection and quantification of oligonucleotides in biological samples, especially tissues, has been limited due to interference from components of the biological matrix and the incompatibility of chromatographic techniques used in the separation of oligonucleotides. For example, attempts to quantify oligonucleotides using mass spectrometry after extraction by solid-phase extraction (SPE) or phenol/chloroform liquid-liquid extraction have shown low recovery and high interference from the biological matrix.²⁶ Various factors such as pH, cation concentration, and organic solvent content can affect the ionization and sensitivity of the detection of oligonucleotides. The use of ion-pair agents and a strong base in the mobile phase has been shown to suppress adduct formation and enhance the sensitivity of ESI-MS.²⁷

Using hypoxanthine phosphoribosyltransferase 1 (HPRT1) gene as an example, here, we report an analytical method for the quantification of the sense and antisense strands of oligonucleotides in mouse liver using multi-mode SPE purification and UHPLC-ESI-MS. This method was successfully applied to a DsiRNA PK study in mouse liver. Further application using this approach to preclinical animal studies will shed light on the efficiency of DsiRNA delivery and the mechanism of RNA silencing.

Materials and Methods

Materials and Chemicals

Female mice (mouse strain: CD 1[cluster of differentiation 1]) were supplied by Harlan Laboratories (Indianapolis, IN). HPRT1 sense and antisense oligonucleotides and internal standard (IS)-1 oligonucleotide (a chemically modified oligonucleotide used as an IS) were obtained from Integrated Deoxyribonucleic Acid Technologies (Coralville, IA) (See Table 1 for sequences). 1,1,1,3,3,3-Hexafluoro-2-propanol (HFIP ≥99%) was obtained from TCI America (Portland, OR). Triethylamine (TEA ≥99%) was obtained from Sigma-Aldrich (St. Louis, MO). Methanol, acetonitrile, and water (HPLC grade) were obtained from J.T. Baker (Phillipsburg, NJ). Clarity OTX 100 mg/sorbent 96 well cartridges and sample loading buffer were obtained from Phenomenex (Torrance, CA).

Table 1.

HPRT1 Duplex and Internal Standards Used in the Study

Strand type	Sequence	Molecular weight
HPRT1 sense	rGrCrCrArGrArCrUrUrUrGrUrUrGrGrArUrUrUrGrArArAdTdT	8037.69
HPRT1 antisense	rArArUrUrUrCrArAmArUmCrCmArAmCrAmArAmGrUmCrUmGrGmCmUmU	8700.35
IS-1	dCdAdGdGdCfUdAfCmGdCdGdTmGdAmGdAdCdAfUfCmAdGdTdAdTfCfCfUmGiT	9440.85

IS, internal standard; HPRT1, hypoxanthine phosphoribosyltransferase 1.

For PK study, the HPRT1 DsiRNA was formulated in stable nucleic acid lipid particles²⁸ and administered at a concentration of 1 mg of DsiRNA/mL of phosphate-buffered saline.

The UHPLC was a Waters Acquity ultra-high-performance liquid chromatography (LC) system equipped with BEH C₁₈, 1.7 μm, and a 2.1 × 50 mm column, and the mass spectrometer was a Waters (Micromass) Quattro Premier XE mass spectrometer (Waters, Milford, MA).

PK Study

Six- to eight-week-old female CD1 mice (approximately 20–25 g) were used. The mice were held in a pathogen-free environment, and all procedures involving animals were performed in accordance with local, state, and federal regulations as applicable and approved by the Institutional Animal Care and Use Committee. The HPRT1 DsiRNA lipid nanoparticles (LNP) was intravenously administered via the lateral tail vein at a dose level of 10 mg/kg in a total volume of 0.01 mL/g body weight (n=3/group, n=1 for last two time points).

The animals were sacrificed by CO₂ asphyxiation at time points of 15 min, 30 min, 1 h, 2 h, 8 h, 24 h, and 72 h. The tissue samples were incubated in Ribonucleic Acid later solution (Ambion, Austin, TX) for 5 min at room temperature and then 4°C overnight before storage at −80°C.

Sample Preparation

Approximately 20 mg to 40 mg liver tissue was homogenized in 100 μL of TEKnova denaturing buffer (TE buffer), pH 7.2 (TEKnova, Hollister, CA) using TissueLyser II (Qiagen, Valencia, CA) with a stainless steel ball at room temperature thrice, for 2 min each, at 26 Hz. The IS-1 (see Table 1) was spiked into each of the homogenized samples to a final concentration of 200 ng/mg to serve as an IS.

Standard calibration curve was prepared in normal untreated female mouse liver homogenate. For this sample, approximately 300 mg liver tissue was homogenized in 1 mL of TE buffer, pH 7.2, using a scaled-up version of the method just described. The sample was then aliquoted into ten 1.7 mL microcentrifuge tubes, and HPRT1 DsiRNA duplex was added to each tube to final concentrations of 1,000, 500, 100, 50, 10, 5, 1, and 0 ng/mg. The HPRT1 DsiRNA at the concentrations of 250 and 25 ng/mg in mouse liver homogenate were separately prepared as quality control (QC) samples. The IS-1 was added to a final concentration of 200 ng/mg in each aliquot.

Oligonucleotide Extraction

To extract oligonucleotides from mouse liver homogenate, the 96-well Clarity OTX cartridge plates (Phenomenex, Torrance, CA) were used with a 96-well plate vacuum manifold (Pall Corporation, Port Washington, NY) following the manufacturer's instructions. Briefly, the 96-well Clarity OTX cartridge plate was conditioned with methanol (1 mL/cartridge) with a vacuum setting of ∼2 mmHg. The 96-well cartridge plate was then equilibrated with equilibration buffer (50 mM NaH₂PO₄ pH 5.5, and 1 mM NaN₃; 1 mL/cartridge) at a vacuum setting of ∼3 mmHg. The mouse liver homogenate samples treated with loading buffer were loaded into the 96-well cartridge plate at a vacuum setting of ∼10–15 mmHg. The sample solutions were completely evacuated through the wells. The 96-well cartridge plate was then washed with 1 mL/cartridge of wash buffer (50 mM NaH₂PO₄, 50% acetonitrile pH 5.5) thrice at a vacuum setting of ∼3 mmHg. The vacuum was then increased to 15 mmHg for ∼30 s before oligonucleotides were eluted initially with 0.5 mL elution buffer (0.1 M NH₄HCO₃/50% methanol pH 8.8) and then with 0.2 mL elution buffer at a vacuum of ∼3 mmHg. The elutes were then combined and dried under nitrogen gas. Finally, the eluted samples were reconstituted in 200 μL of 7 mM TEA (aq)/methanol (9:1, v/v) solution.

LC/MS analysis

Data acquisition and processing for the UHPLC system were performed using MassLynx software. A BEH C₁₈ reversed phase column (1.7 μm, 2.1 × 50 mm) was used. Column temperature was maintained at 65°C. The mobile phases used were as follows: A, 8.6 mM TEA and 100 mM HFIP in HPLC water, pH 8.6; and B, 8.6 mM TEA and 100 mM HFIP in methanol, pH 8.6. The mobile phase B at a gradient from 5% to 22% in a time period of 8 min was used. The sample injection volume was 10 μL, and the flow rate was maintained at 0.2 mL/min.

The temperatures of the ion source and desolvation for the quadrupole mass spectrometer instrument were set at 125°C and 400°C, respectively, and the de-solvation gas flow rate was set at 500 L/h. The electrospray needle voltage was set at 3.0 kV. The HPRT1 sense strand, antisense strand, and IS-1 were scanned from mass-to-charge ratio (m/z) 600 to m/z 2,400 with negative ionization mode. Figure 1a, c, and e are the mass spectra for the HPRT1 sense strand, antisense strand, and IS-1, respectively. Figure 1b, d, and f show the deconvoluted spectra of the HPRT1 sense strand, antisense strand, and IS-1, respectively. The experimental molecular weights obtained by deconvolution of mass spectra were 8,035.7 Da (Dalton) for the sense strand, 8,698.4 Da for the antisense strand, and 9,439.8 Da for the IS-1.

Fig. 1.

The mass spectra of HPRT1 duplex and IS-1 ([predicted m/z]/[experimental m/z]): (a) HPRT1 sense strand (2008.42/2008.25), (c) HPRT1 antisense strand (2174.09/2173.95), and (e) IS-1 (2359.21/2359.10). The deconvoluted spectra by using ProMass deconvolution software for MassLynx (predicted MW [Da]/experimental MW [Da]) : (b) HPRT1 sense strand (8037.69/8035.7), (d) antisense strand (8700.35/8698.4), and (f) IS-1 (9440.85/9439.8). HPRT1, hypoxanthine phosphoribosyltransferase 1; IS, internal standard; Da, Dalton; m/z, mass to charge ratio; MW, molecular weight.

Selected reaction monitoring (SRM) with negative electrospray ionization mode was employed as the analytical method. The −4 charge state ions were selected as SRM parent ions for all three strands based on their signal abundance and ion peak stability. The SRM ion transitions monitored for the MS/MS operation were m/z 2,008.25–1,970.3 for the sense strand, m/z 2,173.95–2,140 for the antisense strand, and m/z 2,359.25–2,325.54 for IS-1. The SRM product ions were established based on the most stable and abundant product ion for each oligonucleotide.

Results and Discussion

UHPLC Separation of the Sense and the Antisense Strands

The HPRT1 DsiRNA sense and antisense strands were separated by reversed-phase (RP) UHPLC using HFIP/TEA buffer system with methanol as an organic eluting solvent. The sense and antisense strands eluted at 5.56 and 6.73 min, respectively (Fig 2a, b). Different combinations of mobile phases that were evaluated for RP-UHPLC to separate oligonucleotides included (i) Mobile phase A: 15 mM TEA, 400 mM HFIP in water; mobile phase B: 8 mM TEA, 200 mM HFIP in methanol, (ii) mobile phase A: 1.7 mM TEA, 400 mM HFIP in water; mobile phase B: methanol, and (iii) mobile phase A: 8.6 mM TEA, 100 mM HFIP in water; mobile phase B: 8.6 mM TEA, 100 mM HFIP in methanol. The first combination of mobile phases did not produce clear and sensitive multiple charge distribution profile, and extensive sodium adduction was observed (data not shown). Figure 3 shows the chromatograms of mobile phases (ii) (Fig. 3c) and (iii) (Fig. 3a, b) at 260 nm wavelength. Even though both mobile phase combination (ii) and (iii) provided baseline separation, the UV responses of all three components in mobile phase combination (iii) (Fig. 3a) were improved by at least 50% when compared with mobile phase combination (ii) (Fig. 3c). Combination (iii) also demonstrated better resolution (Fig. 3a) compared with mobile phase (ii) system (Fig. 3c). We also compared a conventional HPLC column with UHPLC column. Shown in Figure 3 are the chromatograms of UHPLC C₁₈ (particle size 1.7 μm, 2.1 × 50 mm) (Fig. 3a) and conventional C₁₈ (particle size 3.5 μm, 2.1 × 50 mm) (Fig. 3b) with a UV detection of 260 nm. The three components showed 7.5%–75% improvement by the UHPLC column compared with the HPLC column.

Fig. 2.

Chromatographic separation of (a) sense strand, retention time (RT) 5.56 min, and (b) antisense strand, RT 6.73 min.

Fig. 3.

Chromatograms of HPRT1 sense strand, antisense strand, and internal standard (IS-1) with: (a) UHPLC column (C₁₈, 1.7 μm, 2.1 × 50 mm), Mobile phase A: 8.6 mM TEA and 100 mM HFIP in H₂O; Mobile phase B: 8.6 mM TEA and 100 mM HFIP in methanol; (b) HPLC column (C₁₈, 3.5 μm, 2.1 × 50 mm), Mobile phase A: 8.6 mM TEA and 100 mM HFIP in H₂O; Mobile phase B: 8.6 mM TEA and 100 mM HFIP in methanol; and (c) UHPLC column (C₁₈, 1.7 μm, 2.1 × 50 mm), Mobile phase A: 1.7 mM TEA and 400 mM HFIP in H₂O; Mobile phase B: methanol. HFIP, 1,1,1,3,3,3-hexafluoro-2-propanol; TEA, triethylamine; HPLC, high performance liquid chromatography; UHPLC, ultra high performance liquid chromatography.

Electrospray Ionization Tandem Mass Spectrometry

The mass spectra of HPRT1 sense strand, antisense strand, and IS-1 are shown in Figure 1a, c, and e with the deconvoluted spectra shown in Figure 1b, d, and f, respectively. The −4 charge state ions were selected as parent ions for the SRM mode based on their signal abundance and peak stability. In addition, product ion signal abundance and peak stability were also considered for the selection of parent ions for SRM. To increase the abundance of the parent ion that was selected for SRM detection, electrospray voltages were optimized for all three oligonucleotides. As shown in Figure 4, HPRT1 DsiRNA antisense strand demonstrated the improved abundance for the −4 charge state ion after optimization (Fig. 4b) when compared with the original method (Fig. 4a). The loss of a base unit was selected for all three product ions for the SRM mode with the HPRT1 sense minus Guanine and HPRT1 antisense and IS-1 minus Adenine.

Fig. 4.

(a, b) Comparison of different electrospray ionization (ESI) conditions to optimize the −4 charge state ion response. The ESI conditions in (b) were selected for use in the method.

Limited levels of sodium and low levels of potassium adduct ions are observed in mass spectra for all three oligonucleotides. Figure 5 is an enlargement of the −4 charge state HPRT1 sense strand ion peak shown in Figure 1a, revealing the sodium and potassium adduct peaks. The sensitivity of the SRM analysis of the oligonucleotides was inversely proportional to the amount of sodium and potassium adducts. The use of the organic base such as TEA limits the formation of adducts as demonstrated in Figure 5, through displacement of sodium by pairing with the negatively charged phosphate groups of the oligonucleotides, thus reducing the charge states observed in the MS.

Fig. 5.

An enlargement of the HPRT1 sense strand −4 charge state ion peak from Fig. 1a.

Method Validation

Table 2a and b list the standard calibration data for the HPRT1 sense and antisense strands, respectively. The third column is the standard response area that corresponds to the concentration listed in the second column. The IS area corresponds to the IS concentration that was spiked in each standard. The MS response is calculated as (standard response area) × (IS concentration)/(IS response area). The concentration of the analyte was calculated using the equation y=mx + y ₀, where y is the MS response, m represents the sensitivity (slope), x is the analyte concentration, and y ₀ is a constant that describes the background (intercept). The relative deviation (%Dev) on the right column describes how close the calculated value to the mean over the whole calibration range. The calculations were performed with Waters Corporation QuanLynx program. The calibration curves are with weighting factor 1/x ² for more accurate calculation of lower-concentration samples.²⁹

Table 2a.

Hypoxanthine Phosphoribosyltransferase 1 Sense Standard Calibration Data, Prepared Using Untreated Mouse Liver for the In Vivo Study

Type	Std.conc (ng/mg)	Area	IS area	MS response	Detected (ng/mg)	%Dev
Standard	1	36.41	4369.83	1.67	3.69	269.0
Standard	5	84.18	4358.68	3.86	5.02	0.4
Standard	10	195.15	3924.45	9.95	9.94	−0.6
Standard	50	949.69	3130.56	60.67	50.98	2.0
Standard	100	811.79	1436.17	113.05	93.36	−6.6
Standard	500	7375.50	2246.50	656.62	533.15	6.6
Standard	1000	27962.17	4613.49	1212.19	982.65	−1.7
QC	25	479.18	3912.54	24.50	21.71	−13.2
QC	250	3048.10	1747.65	348.82	284.12	13.6

MS, mass spectrometry; QC, Quality control.

Table 2b.

Hypoxanthine Phosphoribosyltransferase 1 Antisense Standard Calibration Data, Prepared Using Untreated Mouse Liver for the In Vivo Study

Type	Std.conc (ng/mg)	Area	IS area	MS response	Detected (ng/mg)	%Dev
Standard	1	42.18	4369.83	1.93	1.03	3.3
Standard	5	148.00	4358.68	6.79	4.41	−11.8
Standard	10	264.18	3924.45	13.46	9.05	−9.5
Standard	50	1062.19	3130.56	67.86	46.85	−6.3
Standard	100	1172.96	1436.17	163.35	113.22	13.2
Standard	500	9187.73	2246.50	817.96	568.18	13.6
Standard	1000	32359.05	4613.49	1402.80	974.64	−2.5
QC	25	634.10	3912.54	32.46	22.25	−11
QC	250	2856.23	1747.65	326.86	226.86	−9.3

The standard curves established from the calibration data (Fig. 6) enabled us to quantify the sense and antisense strands in the mouse liver matrix over a range of 5 ng/mg–1,000 ng/mg and 1 ng/mg–1,000 ng/mg, respectively. The lower limit of detection (LLOD) for the sense strand was ∼1 ng/mg, as a deviation of 1 ng/mg level was more than 20%.²⁵ Both sense and antisense strands showed similar signal intensity. Figure 2a, b show the signal intensity (SRM chromatograms) with samples consisting 0, 1, and 5 ng/mg HPRT1 DsiRNA duplex spiked in liver tissues. The selectivity of the method was demonstrated by the liver sample without oligonucleotides, as shown in Figure 2a, b. There were no detectable matrix interference for HPRT1 sense and antisense strands. Before LC-MS-MS analysis, liver samples were processed by using Phenomenex Clarity OTX cartridge (96-well). Multiple injections (n=3) were made for spiked QC liver samples. There were two precision levels measured for the method; the first precision level was multiple injections (n=3) of the same QC liver samples. The precision was <15% relative standard deviation (RSD) for both QC 1 (25 ng/mg) and QC 2 (250 ng/mg) samples. The second level of precision was the RSDs of liver standards over the linear curve range; this precision was <14% RSD over the standard calibration curve. The data are shown in Table 2a and b for the sense and antisense strands, respectively.

Fig. 6.

Standard calibration curve of HPRT1 antisense (♦; – –) and sense (□ — —) strands in liver tissue. HPRT1 DsiRNA was added to the homogenized liver tissue of 1, 5, 10, 25 (QC1), 50, 100, 250 (QC2), 500, and 1,000 ng/mg, and extracted by SPE. SRM response area was plotted against the oligonucleotide concentration in the liver tissue. SPE, solid phase extraction; DsiRNA, dicer substrate interfering ribonucleic acid; SRM, selected reaction monitoring.

The accuracy of this method is expressed by both the linearity of the spiked standard calibration curve and RSD over the standard calibration range. The correlation coefficient (R ²) over linear calibration curves were 0.9972 (six points) for the sense strand and 0.9931 (seven points) for the antisense strand. The RSDs over the calibration curves were between±0.4% and±13.6% for the sense strand and between±2.5% and±13.6% for the antisense strand. The assay showed no detectable matrix interference for the lower limit of detection at 1 ng/mg and over the calibration curve range. The use of sample matrix in preparing the calibration standards minimized the effects of interference by the matrix in the experimental samples.

Pharmacokinetic Study

The HPRT1 DsiRNA was administered to mice via an intravenous injection, and the oligonucleotide concentration in the liver was quantified by UHPLC-ESI-MS-MS over a 72 h time period. Individual values of liver tissue concentration for each oligonucleotide, as well as the mean values from 3 mice per time point, are reported in Table 3a and b and in Figure 7. Both sense and antisense strands could be detected and quantified as early as 15 min. after injection. The maximum level for both sense and antisense strands of HPRT1 DsiRNA were observed one hour after the injection. After 24 h, the level of both strands decreased below the Lower limits of quantifications (LLOQs) of the method. During the course of 24 h, the sense and antisense strands showed parallel biphasic decay.

Fig. 7.

Linear plot of HPRT1 sense and antisense strand concentration with regard to time in mice liver tissue. The test article was administered in mice intravenously at a dose level of 10 mg/kg in a total vol. of 0.1 mL/10g body weight (N=3/group) at a concentration of 1 mg/mL as a single dose. The oligonucleotide concentration was measured at 15 min, 30 min, 1 h, 2 h, 8 h, 24 h, and 72 h post injection, by UHPLC-ESI-MS. Inset: Log plot of HPRT1 sense and antisense strand concentration with regard to time up to 8 h time point in mice liver tissue. Error bars represent the standard deviation of DsiRNA concentration values. ESI-MS, electrospray ionization tandem mass spectrometry

Table 3a.

Hypoxanthine Phosphoribosyltransferase 1 Dicer Substrate Interfering Ribonucleic Acid Sense Strand Quantification From the In Vivo Study

Time (min)	Mouse ID	Liver weight(mg)	Antisense conc. (ng/mg, liver)	Average (ng/mg)	STDEV
0	1	11.70	<LLOD	NA	NA
15	1	28.30	16.98	17.83	4.81
	2	41.48	13.50
	3	24.10	23.01
30	1	32.10	31.45	33.05	4.73
	2	27.20	38.37
	3	36.00	29.32
60	1	23.10	24.94	25.54	3.40
	2	19.80	22.48
	3	19.20	29.21
120	1	24.70	16.37	16.09	0.74
	2	31.42	15.26
	3	20.80	16.65
480	1	23.60	5.63	4.89	0.85
	2	34.90	3.96
	3	32.00	5.09
1440	1	8.90	<LLOD	NA	NA
4320	1	20.80	<LLOD	NA	NA

LLOD=1 ng/mg, liver. n=3 for all time points except for the time zero and last two time points (n=1).

NA, not available; LLOD, lower limits of detection.

Table 3b.

Hypoxanthine Phosphoribosyltransferase 1 Dicer Substrate Interfering Ribonucleic Acid Antisense Strand Quantification from the In Vivo Study

Time (min)	Mouse ID	Liver weight(mg)	Antisense conc. (ng/mg, liver)	Average (ng/mg)	STDEV
0	1	11.70	<LLOQ	NA	NA
15	1	28.30	17.94	18.98	5.34
	2	41.48	14.24
	3	24.10	24.77
30	1	32.10	36.11	39.33	7.37
	2	27.20	47.77
	3	36.00	34.12
60	1	23.10	32.92	31.07	4.69
	2	19.80	25.73
	3	19.20	34.55
120	1	24.70	18.96	17.55	1.26
	2	31.42	17.16
	3	20.80	16.53
480	1	23.60	4.75	4.26	0.70
	2	34.90	3.45
	3	32.00	4.57
1440	1	8.90	<LLOQ	NA	NA
4320	1	20.80	<LLOQ	NA	NA

LLOQ=1 ng/mg, liver. n=3 for all time points except for the time zero and last two time points (n=1).

LLOQ, lower limits of quantification.

Conclusions

A rapid, sensitive, and specific method based on UHPLC coupled to tandem mass spectrometry has been developed to quantify oligonucleotides in liver tissue. In comparison to traditional HPLC-UV,³⁰ the UHPLC-ESI-MS method we developed here, at least 10 times higher in sensitivity and higher specificity was also obtained by using SRM detection. Using DsiRNA directed against the HPTR1 mRNA as an example, we established the assay dynamic ranges for the sense and antisense strands in mouse liver matrices at 5 ng/mg–1,000 ng/mg and 1 ng/mg–1,000 ng/mg, respectively, with R ²>0.99. The LLOQ is determined as 5 ng/mg for the sense strand and 1 ng/mg for the antisense strand. LLODs for both the sense and antisense strands are ∼1 ng/mg. Using this method, we successfully performed PK study of an LNP formulation of a HPRT1 DsiRNA in mouse liver tissue. We found that HPRT1 DsiRNA directed against the HPRT1 mRNA reaches a peak concentration in mouse liver after 1 h of dosing, and is reduced to below the limit of detection after 24 h. This method will also be tested for application to other tissues, plasma, and urine.

Footnotes

Acknowledgments

The authors thank Lou Lamphere and the Dicerna Pharmaceuticals pharmacology group for providing the liver samples and sample information. The authors also thank Brian Rivera and Greg Scott (Phenomenex, Torrance, CA) for discussions regarding sample preparation techniques.

Disclosure Statement

No competing financial interests exist.

Abbreviations

References

Fire

, Xu

, Montgomery

, Kostas

, Driver

, Mello

. Potent and specific genetic interference by double-stranded RNA in Caenorhabditis elegans. Nature, 1998; 391:806–811.

Mello

, Conte

Jr.

Revealing the world of RNA interference. Nature, 2004; 431:338–342.

Hannon

, Rossi

. Unlocking the potential of the human genome with RNA interference. Nature, 2004; 431:371–378.

Zamore

, Tuschi

, Sharp

, Bartel

. RNAi: double-stranded RNA directs the ATP-dependent cleavage of mRNA at 21 to 23 nucleotide intervals. Cell, 2000; 101:25–33.

Bernstein

, Caudy

, Hammond

, Hannon

. Role for a bidentate ribonuclease in the initiation step of RNA interference. Nature, 2001; 409:363–366.

Kim

, Bhelke

, Rose

, Chang

, Choi

, Rossi

. Synthetic dsRNA Dicer substrates enhance RNAi potency and efficacy. Nat Biotechnol, 2005; 23:222–226.

Rose

, Kim

, Amarzguioui

et al. Functional polarity is introduced by Dicer processing of short substrate RNAs. Nucleic Acids Res, 2005; 33:4140–4156.

Halder

, Kamat

, Landen

et al. Focal adhesion kinase targeting using in vivo short interfering RNA delivery in neutral liposomes for ovarian carcinoma therapy. Clin Cancer Res, 2006; 12:4916–4924.

Takeshita

, Minakuchi

, Nagashara

et al. Efficient delivery of small interfering RNA to bone-metastatic tumors by using atelocollagen in vivo. Proc Natl Acad Sci USA, 2005; 102:12177–12182.

10.

Watanabe

, Harada-Shiba , Suzuki

et al. In vivo siRNA delivery with dendritic poly (L-lysine) for the treatment of hypercholesterolemia. Mol Biosyst, 2009; 5:1306–1310.

11.

Sato

, Murase

, Kato

et al. Resolution of liver cirrhosis using vitamin A-coupled liposomes to deliver siRNA against a collagen-specific chaperone. Nat Biotechnol, 2008; 26:431–442.

12.

Bitko

, Musiyenko

, Shulyayeva

, Barik

. Inhibition of respiratory viruses by nasally administered siRNA. Nat Med, 2005; 11:50–55.

13.

Song

, Lee

, Wang

et al. RNA interference targeting Fas protects mice from fulminant hepatitis. Nat Med, 2003; 9:347–351.

14.

Huber

, Krajete

. Analysis of nucleic acids by capillary ion-pair reversed-phase HPLC coupled to negative-ion electrospray ionization mass spectrometry. Anal Chem, 1999; 71:3730–3739.

15.

Chen

, Qian

, Brennan

, Gallo

. Determination of antisense phosphorothioate oligonucleotides, catabolites in biological fluids, tissue extracts using anion exchange high-performance liquid chromatography, capillary gel electrophoresis. J Chromatogr B Biomed Sci Appl, 1997; 692:43–51.

16.

Bigelow

, Chrin

, Mathews

, McCormack

. High-performance liquid chromatographic analysis of phosphorothioate analogues of oligodeoxynucleotides in biological fluids. J Chromatogr, 1990; 533:133–140.

17.

Zhang

, Lin

, Srinivasan

, Kavetskaia

, Duncan

. Strategies for bioanalysis of an oligonucleotide class macromolecule from rat plasma using liquid chromatography-tandem mass spectrometry. Anal Chem, 2007; 79:416–3424.

18.

Arora

, Knapp

, Reddy

, Weller

, Iversen

. Bioavailability and efficacy of antisense morpholino oligomers targeted to c-Myc and cytochrome P-450 3A2 following oral administration in rats. J Pharm Sci, 2002; 91:1009–1018.

19.

Healy

, Lewis

, Kurz

et al. Pharmacokinetics and biodistribution of novel aptamer composition. Pharm Res, 2004; 21:2234–2246.

20.

Reyderman

, Stavchansky

. Determination of single-stranded oligodeoxynucleotides by capillary gel electrophoresis with laser induced fluorescence and on column derivatization. J Chromatogr A, 1996; 755:271–280.

21.

Deverre

, Boutet

, Boquet

, Ezan

, Grass

, Grognet

. A competitive enzyme hybridization assay for plasma determination of phosphodiester and phosphorothioate antisense oligonucleotides. Nucleic Acids Res, 1997; 25:3584–3589.

22.

, Baker

, Chappell

, Geary

, Cheung

, Levin

. Development of an ultrasensitive noncompetitive hybridization-ligation enzyme-linked immunosorbent assay for the determination of phosphorothioate oligodeoxynucleotide in plasma. Anal Biochem, 2002; 304:19–25.

23.

Tremblay

, Oldfield

. Bioanalysis of siRNA and oligonucleotide therapeutics in biological fluids and tissues. Bioanalysis, 2009; 3:595–609.

24.

Liu

W-L

, Stevenson

, Seymour

, Fisher

. Quantification of siRNA using competitive Qpcr. Nucleic Acids Res, 2009; 37:2–11.

25.

Bansal

, DeStefano

. Key elements of bioanalytical method validation for small molecules. AAPS J, 2007; 9:E119–E114

26.

Johnson

, Guo

, Zang

et al. Quantification of raf antisense oligonucleotide (rafAON) in biological matrices by LC-MS/MS to support pharmacokinetics of a liposome-entrapped rafAON formulation. Biomed Chromatogr, 2005; 19:272–278.

27.

Fountain

, Gilar

, Gerbler

. Analysis of native and chemically modified oligonucleotides by tandem ion-pair reversed-phase high-performance liquid chromatography/electrospray ionization mass spectrometry. Rapid Commun Mass Spectrom, 2003; 17:646–653.

28.

Zimmermann

, Lee

, Akinc

et al. RNAi-mediated gene silencing in non-human primates. Nature, 2006; 441:111–114.

29.

Almeida

, Castel-Branco

, Falcão

. Linear regression for calibration lines revisited: weighting schemes for bioanalytical methods. J Chromatogr B, 2002; 774:215–222.

30.

, Geary

, Levin

. Application of novel quantitative bioanalytical methods for pharmacokinetic and pharmacokinetic/pharmacodynamic assessments of antisense oligonucleotides. Curr Opin Drug Discov Develop, 2004; 7:195–203.