Use of Predicted Versus Measured CCS Values from Different Instrument Platforms, and Isomer Separation on the SELECT SERIES Cyclic IMS

Importance of the Topic

The integration of collision cross section (CCS) measurements into biotransformation studies enhances the reliability of metabolite tracking across different platforms, species, and laboratories. CCS, as an analyte-specific physicochemical parameter, offers a robust means to align isomeric metabolites that may present identical mass spectra and variable retention times in liquid chromatography-MS workflows.

Study Objectives and Overview

This application note evaluates the consistency of measured CCS values obtained on two ion mobility–enabled instruments (SELECT SERIES Cyclic IMS and SYNAPT G2-Si QTof) and compares them to CCS values predicted by a machine-learning algorithm (CCSonDemand). In addition, it demonstrates the capability of cyclic ion mobility separation (cIMS) to resolve closely related isomeric metabolites of ranitidine and imipramine through multiple passes in the mobility cell.

Methodology and Instrumentation

Sample preparation involved 26 approved drugs and their metabolites dissolved in DMSO and diluted to 10 µM in acetonitrile/water (1:1).

• Predicted CCS values: Generated by the CCSonDemand machine-learning model.
• Measured CCS values: Acquired via high-definition MSE (HDMSE) in triplicate on two platforms—SELECT SERIES Cyclic IMS (single-pass mode at 100 ms drift time) and SYNAPT G2-Si QTof—using MassLynx for data acquisition and UNIFI for processing.
• Chromatography: UltraPerformance LC methods ranged from high-throughput to extended gradients to emulate discovery and development conditions.
• cIMS separations: Isomer mixtures were infused at 5 µL/min, and mobility resolution was enhanced by up to 25 passes to achieve baseline separation.

Key Results and Discussion

Comparison of measured CCS values across instruments yielded a root-mean-square error (RMSE) of 1.0%, confirming cross-platform consistency. Predicted versus measured CCS values differed by less than ±5% (bias RMSE of 1.7–2.0%), demonstrating the predictive model’s reliability. The ability to forecast CCS differences guided the use of cIMS:

• Ranitidine N-oxide and S-oxide: Predicted CCS difference of ~3 Å2; resolved after four passes in cIMS.
• Imipramine oxygenated isomers: Required up to 25 passes for complete separation, achieving a mobility resolution of 325 (Ω/ΔΩ).

Benefits and Practical Applications

• Enhanced metabolite alignment across studies, species, and facilities using CCS as a stable identifier.
• Improved structural elucidation of isomeric metabolites when MS/MS and retention times are ambiguous.
• Data-driven decision making on the necessity of high-resolution mobility separations based on predicted CCS values.

Future Trends and Opportunities

The fusion of computational CCS prediction with advanced ion mobility platforms is poised to accelerate metabolite identification and pathway mapping. Anticipated developments include broader CCS databases for diverse compound classes, real-time integration of predicted CCS into acquisition software, and higher-order mobility separations in routine workflows.

Conclusion

Measured CCS values on SELECT SERIES Cyclic IMS and SYNAPT G2-Si QTof exhibit excellent agreement, validating the cross-platform robustness of CCS. Machine-learning predictions closely match experimental values, offering a predictive tool for structural assignment and planning of ion mobility separations. The multipass cIMS capability effectively resolves challenging isomeric metabolites, reinforcing CCS as a critical dimension in modern metabolite analysis.

References

1. ICH M3 (R2) Non-clinical Safety Studies for the Conduct of Human Clinical Trials for Pharmaceuticals, EMA, 2013.
2. FDA Guidance for Industry: Safety Testing of Drug Metabolites, FDA, 2020.
3. Broeckling C. et al., J. Am. Soc. Mass Spectrom., 2021, 32, 661–669.
4. Higton D. et al., Anal. Chem., 2021, 93, 7413–7421.
5. Connolly J.F.R.B. et al., J. Am. Soc. Mass Spectrom., 2021, 32, 1976–1986.
6. Holdsworth C. et al., Poster DMDG 2016.
7. Giles K. et al., Anal. Chem., 2019, 91, 8564–8573.