Slalom Chromatography (Part 2): A Novel Technique to Separate Large Nucleic Acids

In Part II, we demonstrate real-world applications of slalom chromatography (SC) in cell and gene therapy workflows and compare the performance of SC against the conventional agarose gel electrophoresis (AGE)-based separation technique. The column used provides a high-resolution, rapid (< 5 min), and reproducible platform for the analysis of large nucleic acids, offering a robust alternative to slow (> 3 h) traditional AGE. SC has been implemented for high-resolution assessment of plasmid linearization and detecting double-stranded ribonucleic acid RNA (dsRNA), with ~40-fold greater sensitivity than AGE. Collectively, these results demonstrate that SC enables a versatile, high-throughput, and highly sensitive characterization of nucleic acids, supporting process development, quality control, and regulatory compliance for messenger (mRNA)-based therapeutics.

Click here to access the first part of the article

Since the COVID-19 pandemic, messenger ribonucleic acid (mRNA) therapeutics have accelerated, expanding into cancer immunotherapy, human immunodeficiency virus (HIV) vaccine development, and treatments for rare infectious diseases.¹ This growth has brought tighter regulatory expectations for quality and safety. In view of this, our research team believed that slalom chromatography (SC) could be a promising analytical method to support the demand for effective nucleic acid–based treatments.

Although the theory underlying SC was developed decades ago, its practical application and commercialization have not been realized. Initial experiments suggested its potential for purifying nucleic acid samples, but the path to industrial-scale implementation had several key knowledge gaps. Our research not only helped overcome the initial challenges in developing a robust, slalom column–based nucleic acid separation technique but also opened the door to applications across a variety of therapeutic processes aimed at developing high-quality mRNA drugs.

The initial tests and underlying theory of this technique have been described in Part 1.^2–7 Herein, we discuss how SC-based approaches can overcome the limitations of traditional analytical methods at several critical stages in mRNA manufacturing and quality control, such as identifying plasmid topology and detecting double-stranded ribonucleic acid (dsRNA). In developing mRNA-based remedies, plasmid integrity analysis is a critical quality checkpoint before in vitro transcription (IVT). A proper linearization of the plasmid enables accurate IVT and prevents the formation of truncated transcripts. Similarly, detecting and quantifying dsRNA impurities after IVT is crucial, as their trace amounts can trigger innate immunity and reduce therapeutic efficacy.⁸ Traditional analytical methods for assessing plasmid topology and dsRNA contamination—such as gel electrophoresis (GE)—are often time-consuming (1–2 h), require large sample amounts, and lack reproducibility and sensitivity for large nucleic acids. Although enzyme-linked immunosorbent assay (ELISA)—a widely used technique for dsRNA detection— has high sensitivity, it is time-intensive (≈8 h), labor-intensive, and not fully automated, requiring multiple manual steps.^9–10

Our developed SC method surmounts the limits of conventional methods. It is capable of isolating large nucleic acids (up to 30 kbp) and separating linear plasmids from supercoiled forms, with resolution up to 4.5 times that of GE in less than 4 min.¹¹ The columns used displayed exceptional reproducibility, stability, and robust lifetime as demonstrated by very low percent relative standard deviations (% RSD) in retention times for over 400 consecutive injections.¹² We have also implemented a two-step analytical workflow using an SC column to identify dsRNA contaminants in IVT samples. In the first step, single-stranded ribonucleic acid (ssRNA) is selectively digested enzymatically while dsRNA remains intact. In the second step, the intact dsRNA is separated from the degradation products of ssRNA chromatographically, enabling sensitive and specific detection of dsRNA impurities. This method provides 40-fold more sensitive detection of the dsRNA in <6 min compared to 60 min for agarose gel-electrophoresis (AGE), as well as precise discrimination between ssRNA and dsRNA.¹³ The versatility of SC makes it a novel separation technique that could replace traditional analytical methods and evolve the rapidly growing nucleic acid–based therapeutics industry.

Experimental Procedures

Sample Preparation
Plasmid Topology Analysis
Two plasmid samples, pBR322 (New England Biolabs – N3033L), and pCMV-Cas9 (CAS9P-1EA; Millipore Sigma) were procured. Linearization of pBR322 and pCMV-Cas9 was performed by the enzyme digestion method. For a detailed procedure on enzyme digestion of pBR322 and pCMV-Cas9, please refer to the reference.¹¹ The enzyme-solution mix was incubated at 37 °C for 1 h. After incubation, the enzyme was inactivated by incubating the digest at 70 °C for 20 min.

Detection of dsRNA Contamination

Model dsRNA was made by annealing sense and antisense single-stranded RNAs from a 3,000-nucleotide stretch of the Cas9 gene. A 3 kb portion of the Cas9 gene using the Cas9 plasmid was used as a source of DNA templates for RNA synthesis. A detailed procedure for generating deoxyribonucleic acid (DNA) templates from the Cas9 gene via polymerase chain reaction, followed by dsRNA synthesis, has been provided in reference.¹³ To a 20 μL solution mix consisting of 100 ng of ssRNA or dsRNA and 100 mM ammonium acetate buffer (pH 9.0), 200U RapiZyme Cusativin^TM RNase was added for digestion. Enzyme levels need to be scaled up based on the amount of ssRNA in the sample. The mixture was gently vortexed, briefly centrifuged, and incubated at 30 °C for 1 h. Enzyme activity was terminated by heating the samples at 75 °C for 15 min.

Liquid Chromatography (LC) System

Chromatography was carried out using a GTxResolve 250 Å Slalom Column, MaxPeak Premier Technology, 2.5 μm, 4.6 × 300 mm (Waters, p/n: 186011441), installed on Acquity™ UPLC™ I-Class Bio (or equivalent) system, equipped with solvent manager, TUV Detector, and sample manager. The column was stored in 10% (v/v) acetonitrile in 90% aqueous buffer containing 25 mM sodium phosphate (pH 7.0) and 100 mM potassium chloride. Column conditioning was performed by ramping the flow rate up to 1.3 mL/min at 0.1 mL increments over 20 min, followed by equilibration for 40 min. 1X TAE buffer (40 mM Tris-acetate, 1 mM EDTA, pH 8.3; Thermo Fisher Scientific, Cat. No. 15558042) was used as a mobile phase. The system was operated at up to 15,000 psi with a flow rate of 1 mL/min. Empower™ 3.0 software was used to control the instrument, acquire data, and perform analysis.

Results and Discussion

Analysis of Plasmid Topology and Gene Insert Verification by Restriction Digest Using Slalom Chromatography (SC)

SC separates nucleic acids by size and topology through shear‑dependent molecular behavior. Large, rigid dsDNA shows greater flow retardation, while flexible or single‑stranded species migrate faster.^14–15 These shear‑driven differences can also be exploited to resolve linear and supercoiled plasmids.

For example, analysis of the supercoiled (S) E. coli pBR322 (untreated with enzyme) and linearized (L) pBR322 plasmid (treated with the EcoRI enzyme), by slalom column revealed that the supercoiled species eluted earlier at approximately 1.142 min (Figure 1a), while the more readily extended linear form was retained longer, leading to their elution at a later time (~1.277 min) (Figure 1b). This separation pattern was consistent even when a mixed sample containing both supercoiled and linearized pBR322 was examined (Figure 1c). Figure 1 also shows the proficiency of this technique to determine the extent of plasmid DNA linearization in <5 min. These results demonstrate that SC provides a rapid, robust approach for quantifying plasmid linearization efficiency, supporting workflows that require accurate preparation of linear DNA templates.

SC was further evaluated for its ability to confirm the successful insertion of a gene of interest, an analysis typically performed using agarose gel electrophoresis followed by double restriction digestion. To demonstrate this feature, we chose pCMV-Cas9 as a model plasmid and performed its double digestion with NheI and XbaI enzymes. Figure 1d reveals the chromatographic profile of the undigested pCMV-Cas9 plasmid, displaying the expected distribution of its native topological forms. Figure 1e shows the chromatogram of the pCMV-Cas9 treated with the NheI—an enzyme which converts the plasmid to its linear form—revealing a distinct shift in retention time compared with the untreated sample. Figure 1f shows the elution profile of pCMV-Cas9 digested with both NheI and XbaI. The cleavage of the pCMV-Cas9 at NheI and XbaI sites released the 4147 bp Cas9 gene sequence, which is confirmed by the appearance of an elution peak at approximately 1.326 min. Although the linear Cas9 fragment (4,147 bp) and the remaining plasmid backbone (~2,896 bp) differ by only 1,259 bp, the slalom column successfully resolved the two products, with a retention-time difference of 0.077 minutes (Figure 1f).

These studies show that SC provided higher sensitivity and resolution than agarose gels, detecting insert and backbone fragments from 40 ng of material, representing a 5-fold improvement. It clearly resolved undigested, singly digested, and fully digested plasmid species, enabling rapid verification of gene insert incorporation. Overall, the method offers rapid characterization of plasmid topologies, a highly sensitive, high‑resolution alternative to traditional agarose gel electrophoresis, supporting workflows that require accurately linearized DNA for IVT and mRNA production.¹¹

Rapid, Highly Sensitive, and Enzymatically Assisted Detection of dsRNA Contaminants Using Slalom Chromatography

The success of SARS‑CoV‑2 mRNA vaccines has accelerated the growth of mRNA therapeutics, while manufacturing challenges—especially inadvertent dsRNA formation—remain critical.^16–17 Even trace dsRNA can activate innate immune receptors and compromise safety; its sensitive detection is essential. Slalom column coupled with enzyme digestion could enable precise, rapid, and highly sensitive discrimination of ssRNA and dsRNA in IVT samples.

To demonstrate this, we employed a slalom column for the analysis of dsRNA and ssRNA samples treated with RNase (RapiZyme Cusativin) (Figure 2). At first, the chromatographic profiles of the undigested 3000 nt ssRNA and dsRNA were measured (Figures 2a and 2c), which revealed a single intact peak. Then we analyzed the samples treated with RNase. The ssRNA was efficiently cleaved by RNase, producing multiple digestion products, as observed in the chromatogram shown in Figure 2b. The appearance of multiple peaks is likely due to partial entry of the cleavage fragments into the stationary phase pores, allowing separation via hydrodynamic chromatography. In contrast, the dsRNA remained completely intact after RNase treatment, maintaining an identical peak profile in the elution spectrum (Figure 2d). Collectively, these results demonstrate that combining single-strand-specific enzymatic digestion with slalom chromatography provides a robust and reliable approach for distinguishing ssRNA from dsRNA.

SC was also found to be more sensitive than conventional agarose gel electrophoresis. For example, successful detection of 3000 nt dsRNA by the previously mentioned method requires approximately 10 ng of sample (Figure 2e). In contrast, a clear visualization of the bands corresponding to dsRNA in GE requires over 400 ng of samples (Figure 2f). These results demonstrate that slalom chromatography provides significantly higher sensitivity and resolution for dsRNA detection than traditional gel-based approaches, enabling rapid, reliable analysis with minimal sample requirements.¹³

Conclusion

Our research demonstrates that SC provides a high‑resolution, reproducible, and high‑throughput platform for the analysis of large nucleic acids, including plasmid DNA and dsRNA impurities relevant to mRNA therapeutic development. While the approach enables effective separation of high‑molecular‑weight nucleic acids, DNA fragments smaller than 3 kbp cannot be resolved by this technique, irrespective of flow rate or applied shear forces. However, this method offers excellent reproducibility across multiple columns and batches, while significantly reducing analysis time to less than 6 min compared to conventional agarose gel electrophoresis. It allowed precise characterization of plasmid isoforms, efficient assessment of linearization prior to in vitro transcription, and confirmation of gene insertions using restriction digests, with a 5-fold improvement in sensitivity over gel-based methods. Furthermore, the combination of selective RapiZyme Cusativin digestion with SC allowed highly sensitive and specific detection of dsRNA contaminants, providing up to 40-fold greater sensitivity than agarose gel electrophoresis and requiring minimal sample input (~10 ng). The reproducible separation and detection of ssRNA and dsRNA highlights the method’s robustness, accuracy, and suitability for routine analytical workflows in process development and quality control of mRNA therapeutics. Overall, SC offers a versatile and reliable platform for large nucleic acid characterization, supporting safe and effective development of DNA- and mRNA-based therapeutics.

Waters, MaxPeak, BEH, GTxResolve, ACQUITY, UPLC, Empower and RapiZyme are trademarks of Waters Technologies Corporation. TURBO DNase is a trademark of Life Technologies Corporation.SYBR is a trademark of Molecular Probes Inc. Monarch is a trademark of New England Biolabs Inc.

References

1. Ige, M. A.; Ren, X.; Yang, Y.; Zhang, H.; Shen, C.; Jiang, Y.; Li, J. mRNA Therapeutics: Transforming Medicine through Innovation in Design, Delivery, and Disease Treatment. Mol. Ther. Nucleic Acids 2025, 36, 102721. DOI: 10.1016/j.omtn.2025.102721
2. Boyes, B. E.; Walker, D. G.; McGeer, P. L. Separation of Large DNA Restriction Fragments on a Size-Exclusion Column by a Nonideal Mechanism. Anal. Biochem. 1988, 170 (1), 127–134. DOI: 10.1016/0003-2697(88)90099-1
3. Hirabayashi, J.; Kasai, K.-I. Slalom Chromatography: A New Size-Dependent Separation Method for DNA. Nucleic Acids Res. Symp. Ser. 1988, 20, 67–68.
4. Hirabayashi, J.; Kasai, K.-I. Size-Dependent Chromatographic Separation of Double-Stranded DNA Which Is Not Based on Gel Permeation Mode. Anal. Biochem. 1989, 178, 336–341. DOI: 10.1016/0003-2697(89)90649-0
5. Hirabayashi, J.; Ito, N.; Noguchi, K.; Kasai, K.-I. Slalom Chromatography: Size-Dependent Chromatographic Separation of Double-Stranded DNA by a Hydrodynamic Phenomenon. Biochemistry 1990, 29, 9515–9521. DOI: 10.1021/bi00493a004
6. Gritti, F.; Wyndham, K. Retention Mechanism in Combined Hydrodynamic and Slalom Chromatography for Analyzing Large Nucleic Acid Biopolymers Relevant to Cell and Gene Therapies. J. Chromatogr. A 2024, 1730, 465075. DOI: 10.1016/j.chroma.2024.465075
7. Gritti, F. Theoretical Predictions To Facilitate Method Development in Slalom Chromatography for the Separation of Large DNA Molecules. J. Chromatogr. A 2024, 1736, 465379. DOI: 10.1016/j.chroma.2024.465379
8. Boman, J.; Marušič, T.; Seravalli, T. V.; Skok, J.; Pettersson, F.; Nemec, K. Š.; Widmark, H.; Sekirnik, R. Quality by Design Approach To Improve Quality and Decrease Cost of in vitro Transcription of mRNA Using Design of Experiments. Biotechnol. Bioeng. 2024, 121, 3415–3427. DOI: 10.1002/bit.28806
9. Li, D.; Wang, Z.; Wang, D.; Zhang, J.; Sun, F.; Wang, S.; Caccuri, S. Lateral Flow Immunoassay for Rapid and Sensitive Detection of dsRNA Contaminants in in vitro-Transcribed mRNA Products. Mol. Ther. Nucleic Acids 2023, 32, 445–453. DOI: 10.1016/j.omtn.2023.04.005
10. Cook, K. S.; Luo, J.; Guttman, A.; Thompson, L. Vaccine Plasmid Topology Monitoring by Capillary Gel Electrophoresis. Curr. Mol. Med. 2020, 20 (10), 798–805. DOI: 10.2174/1566524020666200427110452
11. Vaishnav, J.; Adapelli, B.; Lauber, M. Plasmid Topology and Digest Analysis Using a GTxResolve™ 250 Å Slalom Chromatography Column; Application Note 720008954; Waters Corporation: Milford, MA, 2025.
12. Vaishnav, J.; Sawyer, K.; Gritti, F.; Fasth, M.; Reidy, C.; Xu, M.; Wyndham, K.; Addepalli, B.; Lauber, M. A New Alternative to Gel Electrophoresis: Higher Resolution and Faster Analysis of Large Nucleic Acids by Rigorously Designed GTxResolve™ 250 Å Slalom Columns; Application Note 720008921; Waters Corporation: Milford, MA, 2025.
13. Vaishnav, J.; Adapelli, B.; Lauber, M. Rapid and Sensitive Detection of dsRNA Contaminants in mRNA Using the GTxResolve™ 250 Å Slalom Chromatography Column; Application Note 720009005; Waters Corporation: Milford, MA, 2025.
14. Gritti, F. Ultra-High Pressure Slalom Chromatography: Application to the Characterization of Large DNA and RNA Samples Relevant in Cell and Gene Therapy. J. Chromatogr. A 2024, 1738, 465487. DOI: 10.1016/j.chroma.2024.465487
15. Gritti, F.; Sawyer, K.; Vaishnav, J.; Adapelli, B.; Lauber, M.; Wyndham, K. Retention and Efficiency of a Novel Slalom Chromatography Column: An Alternative to Agarose Gel Electrophoresis for DNA Separation. J. Chromatogr. A 2025, 1761, 466293. DOI: 10.1016/j.chroma.2025.466293
16. Gholamalipour, Y.; Karunanayake Mudiyanselage, A.; Martin, T. C. 3′ End Additions by T7 RNA Polymerase Are RNA Self-Templated, Distributive, and Diverse in Character—RNA-Seq Analyses. Nucleic Acids Res. 2018, 46, 9253–9263. DOI: 10.1093/nar/gky796
17. Ramírez-Tapia, L. E.; Martin, C. T. New Insights into the Mechanism of Initial Transcription: The T7 RNA Polymerase Mutant P266L Transitions to Elongation at Longer RNA Lengths than Wild Type. J. Biol. Chem. 2012, 287, 37352–37361. DOI: 10.1074/jbc.M112.370643

Download the PDF Here