Specific Determination of Oligonucleotide Therapeutics by Dual Ligation Hybridization Assay

G. A. Tremblay, P R. Oldfield and A. J. Bartlett . Charles River Laboratories Preclinical Services Montreal Inc., 22022 Transcanadienne, Senneville, Quebec, Canada H9X 3R3 Patents pending No. US 61/258 046, US 61/327 245

Objective: Hybridization assays quantify oligonucleotidebased therapeutics in various biological matrices such as plasma or tissues. Current methods fail to discriminate the parent compound from 5’ or 3’ (N-x) truncated metabolites. The objective was to develop an assay specific for the parent compound. Method: The method is based on a bi-enzymatic reaction that results in ligation of probes at both ends of the analyte. Results: Full-length, unmodified or phosphorothioate oligonucleotide therapeutics are quantified with the dual ligation immunoassay. We have demonstrated the mechanism and investigated the specificity of the assay for the analyte. Validation parameters were assessed to document suitability of the method. Conclusion: We have developed, what we believe to be the first hybridization-based assay specific for the parent compound. The method is currently being adapted to the determination of oligonucleotide therapeutics by qPCR.

The analyte, s2B was a phosphodiester DNA OGN with the sequence of a published siRNA3. All oligonucleotides were DNA and have 5’-OH and 3’-OH unless otherwise stated. All oligonucleotides were purified by HPLC. s2B: gcctcagcacgtacctctatt s2B 5’ N-1: cctcagcacgtacctctatt s2B 3’ N-1: gcctcagcacgtacctctat s2B template probe: NH2-gaatagcgaaatagaggtacgtgctga ggcggattcacg-NH2 Ligation probe 1: PO4-tcgctattc-[Biotin-TEG] Ligation probe 2: [Digoxigenin]-cgtgaatcc The oligonucleotides were obtained from IDT Inc. The PNK was purchased from New England Biolabs and the T4 DNA ligase was from USB. The anti-DIG HRP was from Roche Inc. The QuantaBlu and HBC Neutravidin-coated plates were from Pierce Inc.
Figure 1: Schematic representation of the dual ligation immunoassay.
Oligonucleotide Analyte: Template Probe: Ligation Probe 1:

Standard curve • • A representative calibration curve obtained with the dual ligation immunoassay in mouse plasma is shown in Figure 2. The curve working range was 0.12 nM (0.8 ng/mL) up to 4.5 nM (28 ng/mL) using a linear regression model. The behavior and sensitivity of the curve are comparable to other hybridization assays. Specificity for the parent compound • In Figure 4, the signal generated at different concentrations for the FLP of the analyte versus the same analyte truncated by 1 nucleotide at either the 5’-end or the 3’-end (5’ N-1 or 3’ N-1 metabolites) are compared. In this instance, a low interference of ≤4% was observed with the 3’ N-1 metabolite for concentrations above 4.5 nM. In some assays, no interference was detected from the 3’-end. No interference was detected for the 5’ N-1 metabolite in all assays. The dual ligation immunoassay is thus considered specific for the FLP . Validation parameters: precision and accuracy, specificity, prozone and dilution linearity • • • The assay performed comparably in mouse, monkey and human plasma (K2EDTA). In Table 1, the intra-assay precision and accuracy demonstrates that the method reproducibly determines the test OGN within 10% accuracy in mouse plasma. The assay was specific for the analyte, within ±25% of the theoretical concentration in five different lots of mouse plasma (Table 2); the dual ligation assay recovers the analyte irrespective of individual lot variations. No interference was observed in all matrices tested (mouse, human, monkey) and there was no detectable prozone effect when tested at a concentration ca. 100-fold the upper limit of quantitation (ULOQ) in mouse plasma. The recovery of dilution is linear over the range analyzed when performed in mouse plasma (Figure 6).
Oligonucleotide Concentration in Intra-Assay Precision & Accuracy (n=3) Mouse Plasma (nM) § Theoretical Measured CV (%) Recovery (%) Low QC 0.375 0.412 9.2 110.0 Mid QC 2.400 2.506 3.5 104.4 High QC 4.500 4.433 13.3 98.5 § Both 5’ and 3’ (N-1) truncated sequences were <LLOQ and therefore undetected.

Dual Ligation qPCR for Quantifying Oligonucleotide Therapeutics on Dried Blood Spots
Primer 1 Ligation Ligation

Template 1

Template 2 Primer 2


Figure 2: Representative calibration curve with the test oligonucleotide in mouse plasma.



Figure 4: Specificity of the dual ligation assay for the parent compound (FLP) evaluated against the 5’ N-1 metabolite and the 3’ N-1 metabolite in mouse plasma.



Table 1: Intra-assay precision and accuracy of the analyte in mouse plasma

Log concentration (picoMolar)

• • • • •


DIA, Washington, DC

Hybridization assays are used to quantify investigational oligonucleotide (OGN) therapeutics at the discovery, preclinical and clinical stages of the drug development process1. They include sandwich, competitive, ligation2 and nuclease-based3 methods. Until now there was no hybridization assay specific for the parent compound or full-length product (FLP). Indeed metabolites short of 1, 2 or more nucleotides over the FLP would be detected. Specific hybridization assays are highly desirable as they lead to accurate pharmacokinetic (PK) and toxicokinetic (TK) profiles in support of OGN drug development. Although antisense compounds may be active when truncated of a few nucleotides4, other classes of compounds such as siRNA may not be as permissive for sustaining RNA interference activity. A specific method to determine the parent compound will assess the FLP with both 5’- and 3’-ends intact. The ligation reaction could be well suited for that purpose, considering the specificity of T4 DNA ligase for fully matched duplexes. However, ligation cannot be directly implemented at the 5’-end of an OGN analyte since OGN therapeutics are not usually phosphorylated. Therefore, it was envisaged that T4 polynucleotide kinase (PNK) could phosphorylate the 5’-end of the OGN analyte prior to the ligation step. The cloned T4 PNK and T4 DNA ligase available on the market are bacteriophage enzymes involved in DNA repair. Although both enzymes are known to work together, they have not been applied in a quantitative method. The dual ligation immunoassay integrates bi-enzymatic processing of the ON analyte with a defined set of probes, and constitutes the first parent compoundspecific quantitative hybridization assay.
Ligation Probe 2:


Phosphorylation PO4

Table 2: Specificity assessment of the analyte in five different lots of mouse plasma

Bi-enzymatic mechanism of detection
• In Figure 3, when the enzymatic reaction is carriedout with either individual enzyme, no signal is detected above the lower limit of quantitation (LLOQ). For a curve to be generated, both T4 PNK and T4 DNA ligase are required. Thus the reaction is occurring as schematized in Figure 1, where a bi-enzymatic reaction takes place. Also, the washing conditions satisfactorily eliminate un-ligated OGNs. Versatility with the phosphorothioate chemistry • The ligation-hybridization assay typically works with a variety of OGN chemistries including DNA, RNA and phosphothioates2 (PS); T4 DNA ligase is permissive to a variety of nucleic acid chemistries. Fully modified or chimeric PS OGNs are reported to inhibit PNK5. Thus we compared the signal generated with the dual ligation immunoassay with the phosphodiester analyte versus a fully phosphorothioated analogue. The dual ligation immunoassay is permissive to PS OGNs (Figure 5). The inflexion at the end of the curve may be due to a limited inhibition of PNK by PS.





Recovery of the analyte Recovery Independent lots Spiked of mouse plasma (0.25 nM) (% theoretical) Lot #1 0.28 111.2 Lot #2 0.27 106.7 Lot #3 0.31 124.0 Lot #4 0.28 113.3 Lot #5 0.28 110.1 Average: 0.28 113.1

Current miRNA/siRNA qPCR methods lack specificity and largely detect metabolites in addition to the parent compound. We report the application of a dual ligation-based qPCR method (DL-qPCR) for the quantitation of oligonucleotide analytes on dried blood spots. Two template probes guide the ligation of the analyte onto two generic adapters at either end of the analyte (A). qPCR is performed post bi-enzymatic processing using adapters. Based on the threshold cycle (Ct) values obtained from the amplification of the spiked analyte (B), a standard curve was constructed with data points within 20% of the theoretical value (C). No amplification products were detected with the N-1 metabolites from the 5’-end or the 3’-end, making the DL-qPCR specific for the parent compound.

• • •

Figure 6: Dilution linearity assessment of the analyte in mouse plasma.
7 6 5 R2 = 0.9997

• • • The dual ligation hybridization assay is specific for the parent test OGN and does not significantly detect metabolites. The method relies on a bi-enzymatic process comprising DNA ligase and polynucleotide kinase from phage T4. With the added advantage of specificity for the FLP , overall the dual ligation hybridization assay performs similarly to the ligation-hybridization assay2; and notably from the perspective of OGN chemistry versatility. Validation parameters including intra-assay precision and accuracy, specificity and interference, dilution linearity and prozone demonstrated the robustness, reproducibility and accuracy of the assay. One application of the dual ligation hybridization assay is for metabolite testing. The assay can be adapted to determine parent as well as shortmers simply by designing the appropriate (N-x) Template Probe complementary to the desired metabolite. We are currently developing a dual ligation-based qPCR method for the quantitation of oligonucleotide therapeutics.

Process • The dual ligation-hybridization assay workflow is depicted on Figure 1. The template probe is fully complementary to the test OGN, in addition to having extensions on either side that are complementary to Ligation Probes 1 and 2. Ligation Probe 1 is modified for immobilization onto a solid support and synthesized with a phosphate moiety for easier ligation. Ligation Probe 2 is labeled for downstream signaling. The biological sample containing the test OGN is mixed with the Template Probe and Ligation Probe 1 followed by denaturation/annealing/immobilization onto a 96-well plate. Bi-enzymatic reaction: PNK, ligase and Ligation Probe 2 for ca. 1h30 min. The plates are washed thoroughly to remove un-ligated products at the 3’ and 5’-end sides of the Template Probe; in order for signaling to occur, both ends of the test OGN must have ligated. Subsequent signaling follows using standard reagents.

Concentration (nM)

4 3 2 1 0 0 0.002 0.004 0.006 Dilution-1 0.008 0.01 0.012

Figure 3: Mechanism of detection of the dual ligation assay, where reliance on PNK and DNA ligase was tested in mouse plasma.


Figure 5: Comparison of the signal generated with phosphorothioate and phosphodiester variations of the test oligonucleotide in mouse plasma.


1. Tremblay, G.A. and Oldfield, P 2009. Bioanalysis of siRNA and oligonucleotide therapeutics in .R. biological fluids and tissues. Bioanalysis. 1(3), 595–609. 2. Baker BF, Yu Z, Leed JM. 2002. Development of an ultrasensitive noncompetitive hybridizationligation enzyme-linked immunosorbent assay for the determination of phosphorothioate oligodeoxynucleotide in plasma. Anal Biochem. 304(1):19-25. 3. Overhoff, M., Wünsche, W. and Sczakiel, G. 2004. Quantitative detection of siRNA and singlestranded oligonucleotides: relationship between uptake and biological activity of siRNA Nucleic Acids Res. Vol. 32, No. 21, e170. 4. Crooke, S.T. and Lebleu, B., ed. 1993. Antisense Research and Application. CRC Press, Taylor & Francis Group, NY, USA. pp 527-528.10. 5. Teasdale, R.M., Matson, S.J., Fisher, E. and Krieg, A.M. 1994. Inhibition of T4 polynucleotide kinase activity by phosphorothioate and chimeric oligodeoxynucleotides. Antisense Res Dev. 4 (4): 295-97.

• •