An author’s perspective: Development of an LC–MS/MS assay for lipid-conjugated siRNA using a stable isotope labelled internal standard (SILIS)
Authors: Ethan Sanford and Guangnong (Sunny) Zhang (Novo Nordisk; Lexington, MA, USA).
Ethan Sanford is a bioanalytical scientist within Therapeutics Discovery (TD) at Novo Nordisk. He obtained his PhD from Cornell University (NY, USA), where he mapped kinase signalling during DNA repair using a combination of quantitative phosphoproteomics and molecular genetics. He has approximately 4 years of experience in bioanalytical method development across a range of modalities and has worked on both the sponsor and CRO side of bioanalysis. Ethan greatly enjoys working as part of a large and diverse R&D organization.
Guangnong (Sunny) Zhang, PhD, is Director of RNAi Bioanalysis & ADME within Therapeutics Discovery at Novo Nordisk (MA, USA). She has 18+ years of experience leading bioanalytical development across discovery through early clinical, with expertise in PK, ADME/metabolite profiling and biomarker assays using LC–MS/MS, LC–HRMS and molecular platforms. Sunny is experienced in CRO/vendor governance (selection, qualification, technology transfer and oversight) and in delivering regulatory-ready data aligned with GLP/GCP and FDA/EMA/ICH expectations. Previously, she led portfolio bioanalysis in Oncology DMPK at AstraZeneca (MA, USA) and held expanding roles in CRO bioanalytical laboratories at Ricerca Biosciences (OH, USA), Medpace (OH, USA), and Charles River Laboratories (MA, USA).
1. Please could you provide a brief summary of your paper?
Ethan: In this study, we evaluated a stable isotope labelled internal standard (SILIS) strategy for LC–MS/MS quantitation of a lipid-conjugated siRNA. Because the siRNA contained too few phosphorothioate linkages to create sufficient mass shift with 34S labelling, we used a deuterated SILIS (42 deuterium atoms) and did not observe a meaningful “deuterium effect” in ion-pairing reverse-phase chromatography. Overall, the method delivered strong quantitative performance in both mouse plasma and tissue homogenates.
Sunny: From a development-strategy standpoint, SILIS helps turn tissue exposure measurements into decision-ready data, even when tissue matrices are variable or hard to standardize. Why this matters: by reducing matrix-driven bias, it increases confidence in PK and biodistribution readouts for conjugated siRNAs when a well-matched analogue internal standard is not available.
2. How does the development of this LC–MS/MS assay address key oligonucleotide therapeutics challenges, such as tissue-specific delivery and cellular uptake?
Ethan: For me, the link is straightforward: delivery and uptake questions are only answerable with robust quantitation. To understand where an oligonucleotide goes in the body, for what period of time it stays there, and how it gets metabolized, we require assays that reliably measure intact analyte in relevant biological matrices (plasma and tissues in the case of this work). We also want to control for important bioanalytical parameters like extraction recovery and matrix effects.
Sunny: Strategically, the value is in turning complex delivery biology into decision-ready exposure data. As conjugation chemistries diversify, a SILIS-based approach helps create a more platform-like assay strategy that can be applied across programs, reducing the risk that quantitation artefacts—not biology—drive conclusions on tissue delivery and uptake.
3. Why did you choose a stable isotope-labelled internal standard (SILIS) for quantitation, and what advantages did it offer over traditional internal standards in this context?
Ethan: We went with a SILIS because it tracks the analyte as closely as possible during extraction and LC–MS/MS analysis. That matters most when you move between matrices (plasma and tissue homogenates behave differently). We saw in practice that tissue QCs still quantified well against plasma calibration curves.
Sunny: From a program and portfolio perspective, SILIS is attractive because it improves comparability across matrices, studies, and even across sites by minimizing matrix-dependent bias. That reduces rework during method transfer and validation and strengthens confidence that differences observed across tissues or timepoints reflect true biology rather than internal-standard mismatch.
4. How did you ensure the robustness and reproducibility of the assay across different biological matrices, such as plasma and tissue homogenates?
Ethan: We focused on showing the method behaves consistently for our specific use cases. We qualified it in both plasma and tissue homogenates, then tested whether plasma calibration curves could act as a surrogate for tissue. That simplification saves time and reagents and makes method transfer to external CROs easier as long as it is demonstrated during validation/qualification.
Sunny: We approached robustness as a risk-management exercise: define where variability can enter (matrix differences, extraction, calibration strategy) and then generate evidence that the method performs to predefined acceptance criteria in each intended matrix. Where surrogate calibration is considered, we recommend establishing parallelism up front and treating it as a controlled, data-backed simplification—not an assumption.
5. Looking back on the study, is there anything you would have done differently?
Ethan: If we could repeat this work with additional time and budget, I would broaden the chemistries investigated. We might have looked at different lipid moieties with their associated SILISs, for example. That would give us a clearer read on how transferable the workflow is.
Sunny: If we were optimizing for broader impact, I would invest in demonstrating generalizability—multiple conjugate classes, sequences and concentration ranges—so teams can treat the approach as a scalable playbook rather than a single-molecule case study. We would also prioritize early work to de-risk cost and timelines for SILIS synthesis and to understand where hybridization LC–MS/MS could benefit from SILIS addition without introducing probe-competition bias.
6. How do you envision the findings of this study influencing the design of future siRNA therapeutics?
Ethan: In practice, I see SILIS-based LC–MS/MS being used most for lead or later-stage molecules, where the added cost is outweighed by the value of high-confidence quantitation in both plasma and tissues.
Sunny: The broader influence is on development strategy: more reliable tissue quantitation enables clearer exposure–response relationships and faster iteration on conjugate and formulation design. In practice, SILIS is most compelling when it removes uncertainty in key decisions—candidate selection, dose selection, and translational interpretation—while remaining economically justified for the program stage.
7. Are there any plans to extend this assay to other types of oligonucleotide therapeutics, such as ASOs or aptamers, or to other lipid-conjugated molecules?
Ethan: Yes. Next, we want to test additional conjugated oligonucleotides and matrices to understand how well the SILIS LC–MS/MS approach transfers across chemistries and sample types.
Sunny: Yes, the roadmap is to expand chemistry coverage and define clear applicability boundaries (for example, how metabolites, selectivity requirements, and crosstalk affect usable range) so the method can be applied confidently across modalities. If those criteria are met, the same SILIS-enabled framework should translate well to other conjugated siRNAs, ASOs and aptamers.
Disclaimer: the opinions expressed are solely those of the authors and do not express the views or opinions of Bioanalysis Zone or Taylor & Francis Group.
You can find our full collection of Authors Perspectives here >>>
