(T1430-09-50) Toward the Use of In Vitro Animal Matrix Stress Models to Predict In Vivo Stability of Oligonucleotides

Purpose: Oligonucleotides are nucleic acid polymers with the potential to treat or manage a wide range of diseases, e.g. anti-sense oligonucleotides for treatment of neurodegenerative diseases. However, delivery of oligonucleotides to desired sites of action remains a challenge. Current methods of oligonucleotide delivery include lipid nanoparticle encapsulation, brain-penetrating-peptide conjugation, and antibody conjugation. Regardless of delivery strategy, endogenous nucleases and other biogenic factors may contribute to degradation of oligonucleotides in vivo. Efforts to reduce degradation include chemical modification of oligonucleotides and reducing access of oligonucleotides to nucleases (e.g. through delivery strategy). As it is not efficient to routinely evaluate stability of oligonucleotides in vivo at the discovery stage, we developed and used in vitro stress models.

Methods: Oligonucleotides, either free or encapsulated in various nanoparticle formulations, were incubated in both fresh and frozen animal matrices at 37 °C. The matrices used were plasma, whole blood, and liver homogenate of various animal species, including human, monkey, rat, and mouse. At defined time points, the previous mixtures were flash frozen before subsequence analysis. Oligonucleotides were prepared for LC-MS analysis by first spiking in known amounts of internal standard, followed by phenol-chloroform-water extraction. Degradation of oligonucleotides were monitored through absolute quantitation using MRM mode on a SCIEX QTRAP 5500 in negative ion mode. Antibody-oligonucleotide conjugates were similarly incubated and flash frozen, as above. They were then immunoprecipitated from matrices using generic Fc or specific capture on magnetic beads. Following formic acid elution, the extracted conjugated oligonucleotides were analyzed using intact LC-MS on a Thermo Fisher Scientific Orbitrap Q-Exactive Plus MS.

Results: Absolute concentration of free or nanoparticle-encapsulated oligonucleotides was quantified using MRM mode on LC-MS, through interpolation of a concentration vs intensity ratio (analyte/internal standard) standard curve. The LLOQ was 20 ng/mL in mouse plasma and 50 ng/mL in brain homogenate. The accuracy was within 10% and precision was within 15% CV. The concentration of free oligonucleotides can be expected to decrease with increasing degradation. Indeed, this was what was observed when comparing initial timepoints to later timepoints (e.g 0 vs 24 hours) for free oligonucleotides stressed in various animal matrices (e.g. mouse plasma, brain, liver homogenate). Of note, phosphorothioate replacement of the native phosphate backbone in DNA noticeably decreased degradation of oligonucleotides in vitro. Locked-nucleic-acid replacement of canonic bases also decreased degradation. We attribute the decreased degradation to the decreased affinity of nucleases (in animal matrices) to the modified oligonucleotides. Furthermore, we see evidence that nanoparticle encapsulation afforded improved stability. We observed noticeable differences in extent of degradation based on species of animal and how particular animal matrices were prepared. Fresh plasma and whole blood incubation noticeably resulted in greater degradation than their frozen equivalents. EDTA as anti-coagulant significantly reduced the extent of degradation, as compared to heparin. Antibody-oligonucleotide conjugate stability was determined by comparing degraded conjugated oligonucleotide intensity vs naïve conjugated oligonucleotide, using intact LC-MS. This relative comparison approach was chosen due to lack of degraded standards and has been previously reported for use in monitoring degradation of antibody-drug-conjugates¹. After 24 hours stress in in vitro mouse, rat, monkey whole blood, there was significant degradation of free oligonucleotides. However, compared to free oligonucleotides, there is increased stability after conjugation to antibodies, which we attributed to the ability of the particular conjugation site (e.g. on antibodies) to exclude nucleases. This ability of particular conjugation sites to improve stability through exclusion of enzymes has been well-documented for a similar class of molecules, the traditional antibody-drug-conjugate, recently². Compared to in vitro results, the in vivo results showed much more degradation after IV dosing in mice. By 24 hours, there was no remaining intact oligonucleotide attached to the antibody, which was not due to deconjugation. It was possible to deduce that endonucleases were responsible for the initial degradation of the attached oligonucleotide, followed by possible exonuclease activity.

Conclusion: We developed in vitro stress models using various animal matrices to try to predict in vivo stability. It was observed that fresh matrices degraded free oligonucleotides more rapidly than previously frozen matrices. For blood-based matrices, the presence of EDTA notably reduced the degradation observed. Fresh whole blood stress significantly increased the degradation observed. Overall, the in vitro results for free oligonucleotides produced expected behaviors; that is, it showed protective ability of lipid nanoparticle encapsulation and chemical modifications. Compared to in vivo results, it generally rank ordered oligonucleotides by their efficacy. Unfortunately, in vitro stress in mouse, monkey, and rat whole blood of antibody-oligonucleotide conjugates did not predict their stability in vivo. This is unexpected, since the whole blood model was previously shown to predict in vivo stability of antibody-drug-conjugates³. We theorize this disconnection could be caused by enzymes present in the liver in vivo, which whole blood stress would be capture. Hence, we are experimenting with using liver homogenates or other systems to potentially improve translatability with in vivo stability.

References: 1. Xu, et al. Characterization of the drug-to-antibody ratio distribution for antibody-drug conjugates in plasma/serum. Bioanalysis. 2013, 5(9), 1057-71.
2. Sadowsky, et al. Development of Efficient Chemistry to Generate Site-Specific Disulfide-Linked Protein- and Peptide-Payload Conjugates: Application to THIOMAB Antibody-Drug Conjugates. Bioconjugate Chem. 2017, 28(8), 2086–2098.
3. Fourie-O’Donohue, et al. Improved translation of stability for conjugated antibodies using an in vitro whole blood assay. MAbs. 2020, 12(1), 1715705.

Acknowledgements: Authors work for Genentech.


In Vitro Animal Matrixes Stress Assay Correctly Predicted Increased Stability of Phosphorothioate Bonds vs Naive Phosphodiester Bonds


In Vitro Fresh Whole Blood Stress Predicted Increased Stability of Oligonucleotide When Conjugated to Antibody (vs Free)


In Vivo Mouse Plasma Results Showed More Degradation When Compared to In Vitro Fresh Whole Blood Stress Results (Graphic 2), Which Is Contrary to Previous Literature for ADCs