Markus Hassel,a, Milaim Pepaj,b and Birgit M. Wollmann,b
a OsloMet, Pilestredet 50, 0350 Oslo
b Senter for Psykofarmakologi, Diakonhjemmet sykehus, Forskningsveien 13, 0373 Oslo
Email: Markushassel01@gmail.com
Background In the last decades, the oxysterol 4β-hydroxycholesterol (4β-OHC) have been widely studied as a promising endogenous biomarker for activity assessment of the important drug metabolising enzyme cytochrome P450 (CYP) 3A4.1 This has led to the development of several analytical methods using liquid- or gass-chromatography coupled with mass spectrometric detection to measure 4β-OHC, each with different advantages and limitations.2 Initial results are presented from an LC/MS method in development at Senter for Psykofarmakologi, focusing on the optimization of different LC parameters for separating 4β-OHC from a partially co-eluting compound observed in some of the serum samples.
Method The LC-HRMS system consisted of a Vanquish UHPLC coupled to a QExactive Orbitrap mass spectrometer (Thermo Fisher Scientific, Waltham MA, US), and the analytical column used was a Xbridge BEH C18 2.1 x 75 mm with 2.5 µm particle size (Waters, Milford, MS, US). Serum samples known to contain the co-eluting compound were prepared by saponification, protein precipitation and phospholipid removal prior to analysis. To achieve better separation between 4β-OHC and the interfering peak, three different LC parameters were tested: flow rates (0.4, 0.5 and 0.6 mL/min), column temperatures (30°C and 50°C), and different gradient profiles (0-0.5-3-4 min, 0-0.5-4-5 min, and 0-0.5-5-6 min). After identifying the optimal level for each individual parameter, combinations of two parameters were tested at a time. Finally, all three optimized parameters were combined to determine which variables should be adjusted to further optimize the method. In the initial screening, the resolution was estimated using the valley-height method described by John W. Dolan in the LC Troubleshooting Bible3, while the half-height equation for chromatographic resolution was applied for the peaks obtained from the final testing of all optimized parameters.
Results By comparing the estimated resolutions obtained when adjusting the different parameters individually, 0.6 mL/min (Rest = 1.0), 0-0.5-5-6 min gradient profile (Rest = 1.0), and 30°C column temperature (Rest = 1.1) resulted in increased separation. When evaluating the combinations of two parameter adjustments at a time, the results show that modifying the gradient profile to 0-0.5-5-6 min and column temperature to 30°C resulted in the best resolution (Rh = 1.71), while adjusting all three parameters yielded a slightly lower resolution (Rh = 1.62)
Conclusion Although these results showed that a combination of 0.4 mL/min, 0-0.5-5-6 min gradient profile, and a column temperature of 30°C resulted in the highest resolution, increasing the flow to 0.6 mL/min reduced the analysis time per sample while still providing sufficiently high resolution (Rs>1.5). In the future, we plan to improve the sensitivity of the method by optimizing MS tune parameters and automate the sample preparation procedure by using a pipetting robot for liquid handling.
References
1 Gjestad C et al. Br J Clin Pharmacol. 2016;81(2):269-76.
2 Aubry AF et al. Aaps Journal. 2016;18(5):1056-66.
3 Dolan JW. LC Troubleshooting Bible. Sep 2014 [accessed 14 nov 2025]. Available from: https://lctsbible.com/tsb-pdf/32092014.pdf