(T1230-02-10) Unraveling the Bridging Region of Lenacapavir Atropisomers by 2D-LC

Purpose: Atropisomers, arising from restricted bond rotation, often undergo conformational change during liquid chromatography (LC) analysis.¹ This interconversion results in a non-baseline “bridging area” between isomer peaks, potentially obscuring low-concentration impurities. To ensure the purity of atropisomeric compounds, two-dimensional LC (2D-LC) was employed to resolve potential co-eluting impurities.² Lenacapavir Sodium (LEN), classified as a class 2 atropisomer with half-life (t_1/2) < 2 h, was investigated using 2D-LC.² Understanding the atropisomeric behavior of LEN provides valuable insights for developing a quality control-friendly method to support drug substance assay, impurities, and stability testing. On top of that, 2D-LC enables effective impurity identification and separation by utilizing orthogonal separation mechanisms across two dimensions, facilitating the method development process.³ To demonstrate the potential of using 2D-LC for atropisomer analysis, we investigated the retention time window of LEN under pseudo-orthogonal LC conditions and successfully separated a co-eluting regioisomer in a spiked sample at concentrations as low as 0.05% (w/w).

Methods: Sample preparation: LEN reference material and regioisomer control sample (REG) was prepared by dissolving the solid into 4:6 (v/v) water:acetonitrile (ACN) solution. The spike samples were prepared from stock solutions of LEN and REG and mixed at the designated concentration. Chromatographic conditions: First dimension LC condition I: Waters CSH C18 (1.7 µm, 2.1 × 100 mm) column was used. MPA:15 mM ammonium formate (NH₄HCO₂) at pH 10. MPB: ACN. Flow rate: 0.4 mL/min. Column temperature: 30 °C. UV detection: 290 nm (apply to all conditions). Injection volume: 1 µL. First dimension LC condition II: Waters BEH C18 (1.7 µm, 2.1 × 150 mm) column was used. MPA: 90:10 (v/v) buffer:organic solution. Buffer: 45 mM NH₄HCO₂ at pH 8.25. MPB: organic solution made of 60:25:15 (v/v/v) ACN:methanol (MeOH):isopropanol (IPA). Flow rate: 0.3 mL/min. Column temperature: 30 °C. Injection volume: 6 µL. Second dimension LC condition 1: Waters CSH C18 (1.7 µm, 2.1 × 100 mm) column was used. MPA:10 mM NH₄HCO₂ at pH 8-8.4. MPB: ACN. Flow rate: 0.4 mL/min. Column temperature: 15 °C. Second dimension LC condition 2: Waters CSH Phenyl-hexyl (1.7 µm, 2.1 × 100 mm) column was used. MPA: 15 mM NH₄HCO₂ at pH 10. MPB: 35:25:40 (v/v/v) MeOH:tetrahydrofuran (THF):IPA. Flow rate: 0.3 mL/min. Column temperature: 45 °C. Tandem mass spectrometry (MS²) data was collected at collision energy = 30 eV.

Results: A total of 9 cuts were made consecutively across the retention time window of LEN in the first dimension acquired following first dimension LC condition I (Figure 1A). In the first two cuts where the minor atropisomer (LEN.1) was predominantly observed in the first dimension, the major atropisomer (LEN.2) was also observed along with LEN.1 in the second dimension following the second dimension LC condition 1. This is due to the on-column interconversion on the second-dimension column (Figure 1B). Roughly a 1:1 ratio of LEN.1:LEN.2 was observed among the cuts that contained the bridging area (cuts 3-5), whereas LEN.2 was predominantly observed starting at cut 6 (Figure 1C). Applying the first dimension LC condition II, we observed a peak of the regioisomer eluted at 23.11 min in the first dimension at 1% (w/w) spike concentration, between the atropisomer peaks of LEN at 22.23 and 23.62 min (Figure 2A). Second dimension LC condition 1 was applied in 2D-LC separation and the regioisomer peak was observed around 5.9 min in the second-dimension chromatogram (Figure 2B). To circumvent the potential interference of the bridging area, a second 2D method was developed, forcing the atropisomer peaks to coalesce. Following second dimension LC condition 2, a single discrete peak of the regioisomer was found in the total ion chromatogram (TIC) around 10.5 min, whereas LEN atropisomers were found as one single peak eluting around 11 min (Figure 2C). Both second-dimension methods can detect the regioisomer down to 0.05% (w/w). MS² at collision energy (CE) of 30 eV reflected different intensity ratios for four key ions after fragmentation due to the subtle structural differences between the regioisomer and LEN. For the ions at m/z = 968, 948, 889, and 869, the ratio was 0.13:0.15:0.24:0.48, respectively, for LEN (Figure 3C), while the ratio was 0.27:0.13:0.40:0.21 for the regioisomer (Figure 3B).

Conclusion: 2D-LC was employed to comprehensively study the atropisomeric behavior of LEN under various LC conditions, suggesting a facile interconversion between atropisomers from the first dimension to the second dimension. Building on this observation, we developed a second-dimension method to force the atropisomers to coalesce as one discrete peak, effectively mitigating the potential interference from the bridging area. At a concentration of 0.05% (w/w), the regioisomer was indistinguishable in 1D-LC due to the elevated baseline between the atropisomers of LEN. However, our developed 2D-LC method successfully detected and separated the regioisomer from the bridging area. Additionally, MS² revealed different intensity ratios for the two compounds, providing an additional method to distinguish the regioisomer.

References: (1) LaPlante, S. R.; Edwards, P. J.; Fader, L. D.; Jakalian, A.; Hucke, O. Revealing atropisomer axial chirality in drug discovery. ChemMedChem 2011, 6 (3), 505-513. DOI: 10.1002/cmdc.201000485.
(2) Zheng, J.; Lu, B.; Carr, G.; Mwangi, J.; Wang, K.; Hao, J.; Staiger, K. M.; Kozon, N.; Murray, B. P.; Bashir, M.; et al. Lenacapavir Exhibits Atropisomerism - Mechanistic Pharmacokinetics and Disposition Studies of Lenacapavir Reveal Intestinal Excretion as a Major Clearance Pathway. J Pharmacol Exp Ther 2024, 391 (1), 91-103. DOI: 10.1124/jpet.124.002302.
(3) Wang, Q.; He, B. L.; Shackman, J. G. Measuring atropisomers of BMS-986142 using 2DLC as an enabling technology. J Pharm Biomed Anal 2021, 193, 113730. DOI: 10.1016/j.jpba.2020.113730.

Acknowledgements: Yi Shen, Sarju Adhikari, Gavin Carr, and Alban R. Pereira are employees of Gilead Sciences, Inc. and declare no competing interests. We kindly acknowledge the analytical development and operations team in Gilead Sciences, Inc. for the support. We are also grateful to Andrew Hessler, Liping Meng, Tien Ho and all Gilead personnel who contributed to the method development of LEN.


Figure 1. Heart-cutting 2D-LC analysis of atropisomers distribution across the retention time window of LEN.


Figure 2. Separation of regioisomer peak (REG) which is co-eluting in the bridging area of LEN.


Figure 3. MS2 fragmentation at m/z = 968.1 was conducted for the two distinct peaks of LEN and REG observed in the second dimension for the 0.05% (w/w) spiked sample at CE = 30 eV.