Challenges and Solutions in Oligonucleotide Analysis, Part I: An Overview of Liquid Chromatography Methods and Applications

The recent expansion of pharmaceutical portfolios to include more therapeutic oligonucleotides (ONs) has led to a dramatic increase in research around the use of different chromatographic modes for analysis and purification of these molecules. This has been accompanied by a new set of method development and troubleshooting challenges, particularly for those whose background has been focused on small molecule separations, which have not previously been addressed in any detail in the “LC Troubleshooting” column. Thus, I am thrilled to have Martin Gilar join me for a series of “LC Troubleshooting” installments focused on challenges encountered in LC analyses of ONs. Martin is one of the world’s experts on this topic, and this series of articles should be a rich resource for LC practitioners working in this area. - Dwight Stoll

Applied Biosystems, Inc. (ABI), developed the first practical DNA synthesizer in the early 1980s. This enabled custom syntheses of oligonucleotides (ONs) in the range of 2 to 50 bases in length, and facilitated nucleic acid research in academia and the biotech industry. Further enhancements to synthetic methods were made in the following years, providing researchers with chemically modified oligomers of 100 nucleotides or more. Custom ONs can be also purchased from commercial vendors, either as crude synthetic mixtures, or as purified materials. These services have further accelerated nucleic acid research by removing the need for every researcher to purchase their nucleic acid synthesizer. Analytical scientists are now able to obtain practically any type of ON standards needed for method development including ON ladder standards (that is, a sample containing series of molecules with a common sequence, but varying in length by one nucleotide (Figure 1), representative metabolites, internal standards for quantitation, and so on. Moreover, the widespread availability of synthetic ONs has facilitated the study and development of chromatographic methods in both industry and academic laboratories that have become useful in the analysis of nucleic acid therapeutics. This trend is illustrated by the number of recent publications describing new chromatographic modes and methods targeted to analysis of ONs (1–3). Although this is encouraging, many challenges remain and require further analytical tool development. In this series of “LC Troubleshooting” articles, we will discuss challenges commonly encountered in ON analysis, recommended solutions, as well as some challenges that still need to be addressed through further research and development.

Types of Oligonucleotide/Polynucleotide Therapeutic Molecules

Before we can get into discussing specific separation types and challenges we face, we first need to establish the terminology used in the field and the different types of molecules that are encountered by analysts. By classical definition, nucleic acid oligomers consist of 2 to 100 monomeric units or building blocks. These monomeric units are referred to as nucleotides (nt), which consist of a nitrogenous base (adenine, guanine, cytosine, thymine, or uracil), a sugar (ribose), and a phosphate group. Nucleotides are covalently bonded together to form oligomers through linkages between phosphate groups; the basic chemical structure of ONs is shown in Figure 1. Oligomers longer than 100 monomeric units have been classically referred to as polymers. However, in the ON research community this terminology is not adhered to; many researchers refer to nucleic acids longer than 100 nt as oligonucleotides rather than polynucleotides. In this article we will use the oligonucleotide/polynucleotide terminology interchangeably. For example, the polyadenosine (polyA) tail in mRNA is typically longer than 100 nucleotides, but due to the length distribution (heterogeneity) we refer to polyA tail chains as oligonucleotides.

The first nucleic acid molecules used for therapeutic purposes were antisense oligonucleotides (ASOs), typically 18–25 nt long. ASOs work by interfering with protein expression via hybridization to mRNA followed by degradation of the resulting duplex by RNaseH enzyme. First generation ASO therapeutics used a phosphorothioate (PS) modification of the ON phosphate backbone (that is, replacement of one oxygen with a sulfur in every phosphate group of the backbone shown in Figure 1) to improve the resistance of the therapeutic to nucleases (enzymes that degrade nucleic acids by hydrolyzing their phosphate backbone), thus improving ASO and other nucleic acid molecules in vivo half-lives. Second -generation ASO molecules implemented additional chemical modifications such as methylation of the sugars at the 2’ position of specific nucleotides (2’-OMe). Typically two to four 2’-OMe modified nucleotides are placed at the sequence termini to create so-called “gapmer” ONs where the center portion is DNA that can be cleaved. Several ASOs have been approved and sold as drugs over the last 20 years for treatment of various diseases (4).

After discovery of RNA interference molecules in 1998, short interfering RNA (siRNA) molecules were developed as another therapeutic nucleic acid platform. siRNA is formed from two complementary ONs that form an RNA duplex consisting typically of 21–23 base pairs (bp). siRNAs use modifications developed for ASO molecules (PS and 2’-OMe modifications), among others, however phosphorothioate modifications are used more sparsely in siRNA molecules. Rather than complete thiolation as in ASOs, siRNA ONs have only a few selected phosphate linkages modified, typically at 5’ and/or 3’ termini. Recently developed therapeutic siRNA molecules use 2’-OMe, 2’-fluoro (2’-F), or 2’ methoxyethyl (2’-MOE) modified nucleotides. The motivation for using chemically modified bases is similar to that with ASOs: the modifications improve stability of siRNA in vivo and also improve the hybridization affinity towards the target sequences. Additional modifications of siRNA molecules, such as attaching the N-acetylgalactose (GalNac) chemical moiety to chain termini, are used to facilitate siRNA delivery to the target cells. All of the above modifications make production of therapeutic nucleic acids more complex and challenging for analysis.

Recent additions to the family of therapeutic nucleic acids include microRNA (miRNA, 18-25 nt, duplex), single-guide RNA (sgRNA, ~ 100 nt), transfer RNA (tRNA, 70-100 nt), and messenger RNA (mRNA, 1000-12,000 nt). sgRNA molecules used for gene therapy are critical components of the Clustered Regularly Interspaced Short Palindromic Repeats (CRISPR)-CRISPR-associated protein 9 (Cas9) method for gene editing. sgRNA can be synthesized chemically, including the common ON modifications discussed above. Due to their increased length compared to smaller RNAs, synthesis, purification, and quality control of therapeutic sgRNAs is more challenging than for shorter ONs.

So far, tRNA has not been explored as much as other RNAs for therapeutic purposes, but research is ongoing in this area. Naturally occurring tRNAs are more diverse and more heavily modified than other types of endogenous nucleic acids. This makes the development and analysis of tRNA therapies more complex.

Analysis of mRNAs has recently attracted a surge of attention due to their use in vaccines and other potential therapeutic uses. Functional mRNA requires a 5’-cap, which consists of a modified methylated guanosine connected via triphosphate linkage (PPP) to the first nucleotide at the 5′-terminus of the sequence. Similarly important for a mRNA efficacy is a modification of the 3’-end of the sequence with a so-called polyA tail. This polyA tail consists only of riboadenosine, and is approximately 100–150 nt long.

The coding portion of an mRNA sequence provides the instructions for cellular machinery to produce a protein. The size of the targeted proteins correlates with the length of the corresponding mRNA. Typically mRNAs are between 1000 and 5000 nt long, though they can be much longer at 12,000 nt for self-amplifying mRNA (samRNA). These large molecules pose challenges for all available separation methods, including mass spectrometry (MS). This has led to development of methods that use mRNA cleavage strategies involving enzymes (that is, RNases) that cleave the RNA sequence at specific sites to produce shorter ON fragments that are inherently more amenable to analysis with LC and MS. While the development of such RNA mapping methods is a promising approach, digestion of the parent mRNA sequence generates a complex mixture of ONs, and thus requires development of informatic tools similar to those commonly used for peptide mapping and proteomics.

Separation Methods Used for Oligonucleotide Analysis

In this section, we discuss several separation methods for analysis of ONs. Each technique has strengths and weaknesses, including compatibility with MS detection, which are summarized in Table I.

Nucleic acids are polyanionic, highly hydrophilic molecules. Capillary zone electrophoresis (CZE) can separate ONs based on differences between the sizes and charges of different oligomers and their response to an electric field in an open liquid-filled capillary. However, n/n-1 resolution is limited for molecules shorter than about 20 nt (5). In gel electrophoresis (GE), nucleic acids migrate through the gel matrix and are separated according to the number of nucleotides. Polyacrylamide gel electrophoresis (PAGE) was used in early days of DNA sequencing, later replaced by capillary gel electrophoresis (CGE). Agarose gel electrophoresis (AGE) with wide pores is often used for separation of large nucleic acids (>1000 nt in length). Separation is based on the properties of the sieving matrices that preferentially slow down the migration of the longer nucleic acid chains. Gel electrophoresis (GE) can resolve ssDNA with single nucleotide resolution up to several hundred nucleotides in length. Slab gels are used for purification and isolation of small amounts of nucleic acids. However, detection (which is performed by staining nucleic acids with dyes, followed by optical detection of the dye) of nucleic acid bands on gels, and their manual excision for further study, are laborious and often toxic. Moreover, the separation of short 15–30 nt ONs requires densely crosslinked gels, or high viscosity replaceable sieving gel matrices in CGE. For analysis and purification of ONs in the 15–100 nt range, liquid chromatography has become the preferred separation technique. A major driver for the increased use of LC for nucleic acid analysis is its straightforward coupling with mass spectrometric (MS) detection. Hyphenation of GE techniques with MS detection remains challenging.

Anion-Exchange Chromatography (AEC)

Anion-exchange chromatography has a long history of use for separation and purification of nucleic acids. Separation selectivity decreases with increasing ON length, which makes it more difficult to analyze longer ONs. AEC provides single nucleotide resolution up to 20 to 50 nt, depending on the column packing material and particle size (6); an example is shown in Figure 2. Small particle size and/or non-porous AEC sorbents enable n/n-1 resolution for ONs up to 80-100 nt. AEC is typically used for process-scale purification of large quantities of ONs such as therapeutic ASO and siRNA ONs. Due to the polyanionic nature of nucleic acids, the elution from AEC columns is realized with high concentrations of salts, typically between 100 and 500 mM sodium chloride, or even higher. This makes most AEC separations incompatible with MS detection. However, fractions containing separated molecules can be desalted prior to LC–MS analysis (7).

Conventional Reversed-Phase LC (RPLC) and “Trityl On” Methods

Separation of analytes by conventional reversed phase liquid chromatography (RPLC) is largely driven by differences in analyte hydrophobicity. As mentioned above, nucleic acids are highly charged and water-soluble molecules. This leads to poor retention (and resolution) under RPLC conditions unless so-called ion-pairing (IP) reagents are added into the mobile phase. If no IP reagents are added and traditional C18 or polystyrene-divinyl benzene (PS DVB) sorbents are used, nucleotides and ONs are eluted with low levels of organic solvent in the mobile phase (for example, 1-5% of acetonitrile). Only relatively short (<20 nt) ONs can be resolved; mixtures of longer ONs will be coeluted in a single peak. Conventional RPLC separations have been used for purification of ONs labeled with hydrophobic fluorescent dyes, dually-labeled ONs (for example, TaqMan probes, molecular beacons) or dimethoxytrityl (DMT) protected ONs. The DMT hydrophobic group used as a protecting group in ON synthesis is left attached to the full-length product (FLP) and used during isolation of the target sequence. This so-called “trityl on” or “DMT on” method takes an advantage of strong RPLC retention of DMT group to separate the FLP ON from the deprotected (detritylated) byproducts of synthesis. The detritylated ONs, shorter byproducts of ON synthesis (that is, n-x products), are easily resolved from FLP ONs in RPLC without need for ion-pairing additives in the mobile phase (8).

Ion-Pairing RPLC (IP-RPLC)

Separation of ONs by ion-pairing reversed phase liquid chromatography (IP-RPLC) uses positively charged alkylamines as additives to the mobile phase. The IP reagents are adsorbed on the RPLC sorbent and increase the retention and resolution of ONs via electrostatic interactions. The IP reagents can be classified as “weak” (relatively hydrophilic IP reagents including propylamine, diethylamine, and triethylamine), moderately hydrophobic IP reagents (including diisopropylamine, butylamine) and “strong” IP reagents (hydrophobic amines including dibutylamine, hexylamine, and octylamine). IP-RPLC is a mixed-mode separation, where both hydrophobic and electrostatic interactions contribute to ON retention (9). However, when using strong ion-pairing reagents, a high concentration of organic solvent is required to facilitate ON elution, which leads to suppression of the hydrophobic contribution to retention in IP-RPLC. With hydrophobic ion-pairing systems, the ON separation depends mostly on electrostatic interactions and becomes nearly independent of ON sequence and chemical modification (10). Also, IP-RPLC separations with hydrophobic ion-pairing reagents enhance n/n-1 resolution for long ONs (11). This is illustrated in Figure 3, where the resolution of a homooigonucleotide ladder up to 100 nt is achieved with the hexylammonium acetate hydrophobic strong IP system. Figure 3 represents an impressive, but somewhat idealized separation. For ONs consisting of mixed-nucleotides the spacing between n/n-x is not as regular as shown in Figure 3.

The impact of nucleotide hydrophobicity (increasing in order U~C<G<A<T) on n/n-x separation was noted both in IP-RPLC and in HILIC chromatography (Figure 4).

Hydrophilic Interaction Chromatography (HILIC)

HILIC separations have been used less frequently for separations of nucleic acids compared to IP-RPLC, but recently they have gained attention as a suitable ion-pairing-free alternative for LC–MS analysis of ONs. HILIC also provides a different separation selectivity compared to IP-RPLC, which could be another reason to include HILIC in method development, in addition to IP-RPLCor AEC (12). Elution of analytes in HILIC is realized by increasing the concentration of the aqueous component of the mobile phase (with a corresponding decrease in the fraction of acetonitrile) (2,13). Due to their hydrophilic nature, the ONs are well retained in HILIC (typically more than small molecules) and elute from the column starting at around 60% of aqueous buffer. HILIC separations of ONs are performed most successfully on amide stationary phases (2), but diol, zwitterionic, and other stationary phases have also been used (2,12). It seems likely that electrostatic and hydrogen bonding interactions contribute to retention in addition to a HILIC partitioning mechanism (analyte distribution between water-enriched mobile phase at the sorbent surface and acetonitrile-rich bulk mobile phase). However, no mechanistic investigation of HILIC ON retention has been published yet, as far as we know. HILIC is promising as a technique complementary to IP-RPLC and offers an alternative separation selectivity (see Figure 4 and reference [12]). However, the n/n-1 separation of long (>40 nt) ONs does not match the capabilities of optimized IP-RPLC methods (11).

Size-Exclusion Chromatography (SEC)

Size-exclusion chromatography is a technique that separates molecules based on their hydrodynamic size. Smaller molecules can enter sorbent pores more readily, therefore, they elute later than larger molecules, which are partially or completely excluded from the pores. SEC is frequently used for separation of proteins and polymers. Surprisingly few reports have described application of SEC to separation of nucleic acids. Recent applications include SEC separation of siRNA duplex from its single-stranded constituents (14), and separation of poly(A) tail liberated from an intact mRNA by mRNA digestion (11). Linear ONs and nucleic acids have larger hydrodynamic radii compared to molecules with the same molecular weight, but non-linear shapes. Therefore, they elute earlier in SEC than more compactly folded proteins of a comparable molecular weight. As a guideline, short ONs (2 to 25 nt) are best resolved using 125 Å sorbent, and separation of siRNA duplex from single stranded ONs can be performed using a 200 Å sorbent. Separation of 100–150 nt poly(A) tails from mRNA and short ONs was accomplished using 250 Å sorbent, and mixtures of intact mRNAs in the range of 1000 to 10000 nt were analyzed using SEC columns packed with 450 Å, 1000 Å, or 2000 Å sorbents (15). SEC generally provides lower resolving power compared to other modes of chromatography. As a rule of thumb, SEC separation is possible for molecules that differ by a factor of two in hydrodynamic size or molecular weight. In other words, a 25-nt ON can be well resolved from a 50-nt ON, 1000-nt from 2000-nt, and so on, provided that a SEC sorbent with an appropriate pore size is chosen. Chromatograms illustrating the effect of pore size on SEC separations are shown in Figure 5.

Hydrophobic Interaction Chromatography (HIC)

Hydrophobic interaction chromatography is often used for separation of macromolecules. Opposite from AEC, in HIC separations retention increases with increasing mobile phase salt concentration. Under these conditions analytes adsorb onto a hydrophobic stationary phase (typically phenyl- or C4-modified sorbent), and are retained there until the mobile phase salt concentration is reduced substantially to promote elution. Although commonly used for protein separations, HIC has also been used for separation of nucleic acids from proteins (16), and for separation of ONs according to their hydrophobicity (17). HIC is not compatible with LC–MS due to the high salt concentrations required, and it is rarely used for separation of therapeutic ONs.

Affinity Chromatography

Selective isolation of nucleic acids by sequence specific hybridization with the complementary strand immobilized on the surface of titration well plates, or on magnetic beads, is commonly practiced in nucleic acid research. Selective hybridization is also utilized for isolation of mRNA by chromatography. Deoxythymidine ONs immobilized on POROS particles, on monolithic supports, or PS-DVB particles are commercially available for enrichment of mRNA molecules containing a poly(A) tail (18,19). In this type of separation, nucleic acid sequences and other sample components that do not contain a polyA tail are washed from the column unretained. mRNAs with poly(A) tails hybridize to immobilized complementary oligodeoxythymidine chains under high salt conditions and are eluted with water. Elevated temperature can also be used to destabilize the duplex and elute mRNA sequences containing poly(A) tails.

Affinity chromatography is not limited to isolation of mRNA; it can be extended to other applications by immobilizing a specific ON complementary to a target nucleic acid sequence.

Summary

The recent expansion of pharmaceutical portfolios to include more therapeutic ONs has led to a dramatic increase in research around the use of different chromatographic modes for analysis and purification of these molecules. This has been accompanied by a new set of method development and troubleshooting challenges, particularly for those whose background has been focused on small molecule separations. In this installment, we have reviewed the different types of analytes and samples encountered in this space, as well as the different chromatographic techniques that can be used to address them. IP-RPLC is currently the most flexible and LC–MS compatible method. In the next installment in this series, we will focus more on the IP-RPLC technique specifically, discuss the mechanism of retention of ONs under IP-RPLC conditions, options for selection of ion-pairing reagents and mobile phase conditions, the impact of IP selection on the ON separation, and LC–MS. Finally, we will discuss the impact of ON modification including phosphorothioate backbone modifications on ON separation success.

Gen-Pak, ACQUITY, and BEH are trademarks of Waters Technologies Corporation. TaqMan is a trademark of Roche Molecular Systems, Inc. POROS is a trademark of Applied Biosystems, LLC.

References

1. Donegan, M.; Nguyen, J. M.; Gilar, M. Effect of Ion-Pairing Reagent Hydrophobicity on Liquid Chromatography and Mass Spectrometry Analysis of Oligonucleotides. J. Chromatog. A 2022, 1666, 462860. DOI: 10.1016/j.chroma.2022.462860
2. Lobue, P. A.; Jora, M.; Addepalli, B.; Limbach, P. A. Oligonucleotide Analysis by Hydrophilic Interaction Liquid Chromatography-Mass Spectrometry in the Absence of Ion-Pair Reagents. J. Chromatog. A 2019, 1595, 39–48. DOI: 10.1016/j.chroma.2019.02.016
3. Togawa, H.; Okubo, T.; Horiuchi, K.; Yamaguchi, T.; Tomita-Sudo, E.; Akita, T.; Kawakami, J.; Obika, S. Separation of the Diastereomers of Phosphorothioated siRNAs by Anion-Exchange Chromatography under Non-Denaturing Conditions. J. Chromatog. A 2024, 1721, 464847. DOI: 10.1016/j.chroma.2024.464847
4. Vinjamuri, B. P.; Pan, J.; Peng, P. A Review on Commercial Oligonucleotide Drug Products. J. Pharm. Sci. 2024, 113 (7), 1749–1768. DOI: 10.1016/j.xphs.2024.04.021
5. Demelenne, A.; Gou, M.-J.; Nys, G.; Parulski, C.; Crommen, J.; Servais, A.-C.; Fillet, M. Evaluation of Hydrophilic Interaction Liquid Chromatography, Capillary Zone Electrophoresis and Drift Tube Ion-Mobility Quadrupole Time of Flight Mass Spectrometry for the Characterization of Phosphodiester and Phosphorothioate Oligonucleotides. J. Chromatog. A 2020, 1614, 460716. DOI: 10.1016/j.chroma.2019.460716.
6. Thayer, J. R. High-Resolution Separation of Oligonucleotides on a Pellicular Anion-Exchange Column. Thermo Fisher Scientific Application Note. 2016. https://assets.thermofisher.com/TFS-Assets/CMD/Application-Notes/AN-20996-LC-DNAPac-PA200-Oligonucleotides-AN20996-EN.pdf (accessed 2025-07-27).
7. Thayer, J. R.; Puri, N.; Burnett, C.; Hail, M.; Rao, S. Identification of RNA Linkage Isomers by Anion Exchange Purification with Electrospray Ionization Mass Spectrometry of Automatically Desalted Phosphodiesterase-II Digests. Anal. Biochem. 2010, 399 (1), 110–117. DOI: 10.1016/j.ab.2009.11.009
8. Hill, T. L.; Mayhew, J. W. Convenient Purification of Tritylated and Detritylated Oligonucleotides up to 100-Mer. J. Chromatog. A 1990, 512, 415–431. DOI: 10.1016/S0021-9673(01)89508-X
9. Gilar, M.; Bouvier, E. S. P. Purification of Crude DNA Oligonucleotides by Solid-Phase Extraction and Reversed-Phase High-Performance Liquid Chromatography. J. Chromatog. A 2000, 890 (1), 167–177. DOI: 10.1016/S0021-9673(00)00521-5
10. Gilar, M.; Schomann, N.; Schott, S.; Rühl, M. Impact of Nucleotide Hydrophobicity on Oligonucleotides Separation in Liquid Chromatography. J. Chromatog. A 2025, 1753, 465968. DOI: 10.1016/j.chroma.2025.465968
11. Gilar, M.; Doneanu, C.; Gaye, M. M. Liquid Chromatography Methods for Analysis of mRNA Poly(A) Tail Length and Heterogeneity. Anal. Chem. 2023, 95 (38), 14308–14316. DOI: 10.1021/acs.analchem.3c02552
12. Gilar, M.; Koshel, B. M.; Birdsall, R. E. Ion-Pair Reversed-Phase and Hydrophilic Interaction Chromatography Methods for Analysis of Phosphorothioate Oligonucleotides. J. Chromatog. A 2023, 1712, 464475. DOI: 10.1016/j.chroma.2023.464475
13. Lardeux, H.; Guillarme, D.; D’Atri, V. Comprehensive Evaluation of Zwitterionic Hydrophilic Liquid Chromatography Stationary Phases for Oligonucleotide Characterization. J. Chromatog. A 2023, 1690, 463785. DOI: 10.1016/j.chroma.2023.463785
14. Gilar, M.; Redstone, S.; Gomes, A. Impact of Mobile and Stationary Phases on siRNA Duplex Stability in Liquid Chromatography. J. Chromatog. A 2024, 1733, 465285. DOI: 10.1016/j.chroma.2024.465285
15. Goyon, A.; Tang, S.; Fekete, S.; Nguyen, D.; Hofmann, K.; Wang, S.; Shatz-Binder, W.; Fernandez, K. I.; Hecht, E. S.; Lauber, M.; Zhang, K. Separation of Plasmid DNA Topological Forms, Messenger RNA, and Lipid Nanoparticle Aggregates Using an Ultrawide Pore Size Exclusion Chromatography Column. Anal. Chem. 2023, 95 (40), 15017–15024. DOI: 10.1021/acs.analchem.3c02944
16. Schafer-Nielsen, C.; Rose, C. Separation of Nucleic Acids and Chromatin Proteins by Hydrophobic Interaction Chromatography. BBA - Gene Struct. Expression 1982, 696 (3), 323–331. DOI: 10.1016/0167-4781(82)90064-1
17. Diogo, M. M.; Queiroz, J. A.; Prazeres, D. M. F. Studies on the Retention of Plasmid DNA and Escherichia Coli Nucleic Acids by Hydrophobic Interaction Chromatography. Bioseparation 2001, 10 (4–5), 211–220. DOI: 10.1023/a:1016361721316
18. Dewar, E. A.; Guterstam, P.; Holland, D.; Lindman, S.; Lundbäck, P.; Brito Dos Santos, S.; Wang, S.; Swartz, A. R. Improved mRNA Affinity Chromatography Binding Capacity and Throughput Using an Oligo-dT Immobilized Electrospun Polymer Nanofiber Adsorbent. J. Chromatog. A 2024, 1717, 464670. DOI: 10.1016/j.chroma.2024.464670
19. Grinsted, J.; Liddell, J.; Bouleghlimat, E.; Kwok, K. Y.; Taylor, G.; Marques, M. P. C.; Bracewell, D. G. Purification of Therapeutic & Prophylactic mRNA by Affinity Chromatography. CGTI 2022, 8 (2), 335–349. DOI: 10.18609/cgti.2022.049