Antibody–oligonucleotide conjugates, or AOCs, bring together two biomolecule classes with very different analytical behaviors. Antibodies behave like proteins. Oligonucleotides bring permanent negative charge, secondary structure, and different stability considerations.
That combination creates new challenges for chromatographic method development. In this Q&A, Shilin Cheung, Staff Scientist at Phenomenex, explains why AOCs require analysts to think across both protein and oligonucleotide workflows—and why recovery, sample handling, and communication with manufacturers matter as much as the separation itself.
AOCs bring together antibodies and oligonucleotides, which have very different analytical behaviors. What makes their characterization more complex than standard antibody or oligonucleotide analysis?
The main challenge is that antibodies and oligonucleotides have very different bio physicochemical properties. Traditionally, when we work with antibodies, we treat them as proteins. We think about amino acids, charge, isoelectric point (pI), and protein folding. We usually try to avoid extreme pH or high temperatures because we need to consider protein stability, tertiary structure, and unwanted interactions.
Working in analytical science?
Register for a FREE Separation Science account to subscribe to the Separation Science Newsletter.
A lot of chromatography for proteins is designed around those concerns. In size exclusion chromatography, for example, we often use higher salt and phosphate buffer because proteins can interact with the stationary phase.
Oligonucleotides are very different. Their backbone carries a permanent negative charge. They are generally more robust across temperature and pH conditions. Their structures can be reversible, and higher temperatures can help disrupt secondary structures. That can extend the oligonucleotide and produce sharper peaks, especially in ion-pairing reversed-phase methods.
When you combine an antibody with an oligonucleotide, you have to balance both sets of behaviors. A higher temperature may help the oligonucleotide, but it may also cause the protein to unfold or aggregate, foul the column, clog the system, or create unwanted interactions. A high-salt condition may help a protein separation, but it may not be ideal for an oligonucleotide.
That is why method design becomes so important. You cannot just apply a protein method or an oligonucleotide method and assume it will work for the conjugate.
Where do size exclusion chromatography and ion exchange separations add the most value? What information can each technique reveal?
For biologics, people often go in one of two directions. They either keep the molecule under native-friendly conditions, or they fully denature it and start fragmenting it for deeper characterization.
Usually when I work on a new biological modality, I prefer to use native LC methods first to gain a deeper understanding of how best to handle these analytes. For AOCs, techniques such as ion exchange and size exclusion lend themselves well to that approach. They use salt concentrations and relatively neutral pH conditions that can make it easier to inject samples from different stages of production.
Size exclusion chromatography can help you look at size-related forms, such as fragments, monomers, aggregates, or larger species. Ion exchange can help you understand charge-related differences, which may reflect the antibody, the oligonucleotide, or the conjugated forms.
If you are working in a mass spectrometry environment, you may also use enzymes or other approaches to break the molecule down and perform additional characterization. But no matter what detector you use—UV, MALS, CAD, or mass spec—the front-end LC method has to suit the purpose.
You also have to make sure the sample actually gets into the system and through the system. The vial, pipette tip, dilution buffer, mobile phase, LC system, column, and detector can all affect recovery. That is especially important for AOCs because proteins can stick to surfaces, systems, or columns, and the mobile phases used for proteins are often less mass-spec-friendly than those used for oligonucleotides.
How can techniques such as UV, MALS, and mass spectrometry work together to build confidence in AOC characterization?
Because AOCs are still relatively new, a lot of the early analytical work focuses on a basic question: What does the molecule look like in its formulation matrix?
For me, UV is always important. I like to put a UV detector in front because I want to make sure I can see what I injected. Oligonucleotides have strong absorbance at 260 nm. Proteins have characteristic absorbance at 280 nm, and you can also monitor lower wavelengths, such as 220 or 215 nm, to look at the peptide backbone. The ratios between those characteristic wavelengths are also very useful identifiers.
UV is robust because you do not have to worry about ionization, charge states, gas-phase behavior, or other mass spectrometry variables. It gives you a more universal way to see the sample.
MALS can add another layer of information. Multi-angle light scattering uses the scattering properties of the biomolecules. It is very useful for larger biologics, mRNA, lipid nanoparticles, and antibodies, and it can also be useful in AOC work.
The raw MALS signal can tell you when larger species are coming through. If everything is small, the signal may stay flat. But if the biomolecule is aggregating, or if something unexpected is coming off the column, you can see a strong scattering signal.
Other detectors can also help. CAD can provide complementary information. Mass spectrometry can provide identification and structural detail. But every detector has its own strengths and limitations. MS depends on ionization. UV depends on absorbance. MALS responds better to larger species but still relies on some assumptions on the biomolecule.
That is why combining detectors can be useful during early method development. The more feedback you get, the better you can understand the various species observed in the separation.
What are the biggest method-development challenges when trying to preserve conjugate integrity while still achieving meaningful separation?
AOCs are still relatively new. Even the manufacturers making them are still optimizing how to link the antibody and oligonucleotide together. Depending on the chemistry, that linkage may involve lysine, cysteine, or another strategy. The stability of those linkages can be hard to predict. Heat, ionic strength, pH, light, or other conditions may affect the molecule. That means sample handling becomes very important.
These samples are often delivered via a cold chain and analytical development starts at the point of receipt. I usually aliquot the sample and store the extra aliquots in at least two separate temperatures to aid in stability studies and further develop methods to preserve them as much as possible.
Controls are also important. If I am working with an AOC that combines an antibody with an siRNA component, I want to ask the manufacturer whether I can obtain the antibody and the oligonucleotide separately. Then I can run the AOC, the antibody, and the oligonucleotide. That helps me understand whether I am seeing the intact conjugate, leftover reactants, free antibody, free oligonucleotide, or something else.
It is like a puzzle. If A plus B equals C, then when you run C and still see A and B, you need to understand whether those are leftover components, degradation products, or part of the sample profile.
One of the most important things an analytical chemist can do is talk to the manufacturer. The manufacturer understands the production process. They know the pH, buffers, filtration steps, and conditions used to preserve the molecule. If I choose conditions close to what they already know works, the sample has a better chance of surviving the method.
I have seen people buy an AOC or an antibody and then run it in a completely different mobile phase. Sometimes they dilute it in water because they assume water is safe. But I would always ask: Are you sure? Why not talk to the manufacturer first?
Manufacturers often have simple but sturdy analytical methods. They may use UV, gels, BCA assays, or other older methods to check protein and oligonucleotide content. These methods may not be highly quantitative, but they tell the manufacturer whether the components are present and whether the yield is acceptable.
That knowledge is valuable because yield is money for the manufacturer. They know how to preserve it.
How has manufacturer input helped you interpret unexpected results during AOC method development?
In one AOC study, I worked with an antibody conjugated to an siRNA, which is a double-stranded oligonucleotide. I also worked with an antibody conjugated to a single-stranded oligonucleotide. The single-stranded oligonucleotide conjugate was much harder to work with. Its final concentration was lower. When I changed the ionic strength of the mobile phase, the aggregate and monomeric forms changed.
I reached out to the manufacturer and explained what I was seeing. They told me they had experienced a large yield loss in the final step. When they combined the antibody with the single-stranded oligonucleotide, they lost about 75 percent of their yield during one filtration step. The solution went cloudy.
That information was very helpful. Without it, I might have spent a lot of time questioning the vial, the LC system, the detector, the column, or the mobile phase. But the manufacturer already knew this construct was more difficult. That kind of conversation helps you understand whether what you are seeing is a method problem or a sample problem.
For someone moving into AOC analysis, what should they prioritize first: separation robustness, detection strategy, sample handling, or data interpretation?
The first thing I would recommend is to work with the separate analytes and understand the method conditions for both sides of the molecule. You need to understand the protein side. You also need to understand the oligonucleotide side. Learn why certain methods work for each component and what their limitations are.
Ion-pair reversed-phase chromatography is a standard approach for oligonucleotides. But many proteins may not elute under those same conditions. Sometimes people run an AOC using an oligonucleotide method and look only for the oligonucleotide. They may not check whether the antibody came off the column.
That is a problem. You have to keep your eyes open. Whatever analytical method you develop, you need to understand recovery. You need to see what you put in. If you do not see it coming out, then you lost it somewhere in the system. So for me, recovery comes first. Before you interpret the data, ask whether the full sample made it through the workflow.
What is the key message for analytical scientists working with AOCs?
Do not assume that an antibody method or an oligonucleotide method will automatically work for an AOC. When these molecules combine, they may display very different analytical behaviors. The protein may need one set of conditions, while the oligonucleotide may benefit from another. The method has to balance both.
Start with the individual components. Use controls. Talk to the manufacturer. Choose detectors that help you see the full sample. Most importantly, make sure that what you inject is what you recover. If you lose part of the molecule in the system, the separation may look clean, but the method is not telling you the full story.
This interview has been edited for length and clarity.
