The Analysis of Bile Acids: Enhancement of Specificity Using an Ion Mobility-Tofms Based Approach

THE ANALYSIS OF BILE ACIDS: ENHANCEMENT OF SPECIFICITY USING AN ION MOBILITY-TOFMS BASED APPROACH

Jonathan P Williams1, Martin Palmer1, Jonas Abdel-Khalik2, Yuqin Wang2, Sarah M Stow3, Mark Towers1, Giuseppe Astarita1, James Langridge1 and William J Griffiths2 1 Waters Corporation, Wilmslow, Manchester UK; 2 College of Medicine, Swansea University UK; 3 Laboratory for Structural Mass Spectrometry, Vanderbilt University, TN, USA

Fig.1 shows a schematic of Vion. In brief, the instrument com- MALDI Imaging Ion Mobility MS measurements of the 100 Representative conformations from distance geometry INTRODUCTION prises an IM separation device, a quadrupole and segmented RESULTS isomeric bile acids deoxycholic acid and hyodeoxy-cholic modeling for the bile acids investigated collision cell prior to the TOFMS. Ions are accumulated in the acid DCA 'Steroidomics' is the qualitative and quantitative trap travelling-wave (T-Wave) and periodically released into New and improved methods were sought for the identification, HA quantification, and characterization of bile acids, oxysterols, and other

the T-Wave IM where they separate according to their mobil- % study of steroid-type molecules found within the sterols and steroids. The involvement of these molecules in metabolome. Bile acids for example, are ity. neurogenesis and immunity is investigated. classified as acidic sterols that are synthesised The use of IM as an analytical tool to aid direct infusion ESI, DESI and mainly by the liver from cholesterol and aid MALDI shotgun steroidomic-type analysis was investigated. Bile acids 0 digestion and fat solubilisation. The presence of present themselves in biological type samples as complex mixtures. 2.00 2.50 3.00 3.50 4.00 4.50 5.00 multiple isomeric bile acids poses a great Structural information may be obtained using MS/MS but in the absence of a chromatographic step, unambiguous characterisation using MS/MS 100 challenge for steroidomic research. Ion mobility can be challenging since the selected precursor ion can be composed of DCA -mass spectrometry (IM-MS) was combined chemical isomers and interfering isobaric ions. with molecular modelling for the separation and The individual rotationally averaged CCS measured experimentally using configurational analysis of thirteen medically direct infusion ESI-T-Wave ion mobility are shown in Table 1. The results % CDA relevant bile acids. The usefulness of the represent an average of three measurements. Isomeric bile acids are rotationally averaged collision cross-section highlighted in grey.

(CCS) information derived from the experiment- 0 Fig. 1 Schematic of the Vion IMS Q-ToF 2.00 2.50 3.00 3.50 4.00 4.50 5.00 tally derived IM measurements of relevant bile Direct infusion-ESI Ion Mobility MS: CCS measurements acids may be used to enhance specificity and of the bile acids investigated 100 augment steroidomic-type research and aid the IMS comprises a travelling wave RF ion guide, which incorpo- DCA diagnosis, prognosis and management of G2-Si rates a repeating sequence of transient DC pulses to propel - Vion 2 CCSN2 BILE ACID [M-H] CCSN2 (Å ) 2

disease. ions through the guide in the presence of N2 bath gas. Upon (Å ) % exiting the IMS cell, ions can be selected with the quadrupole Fig. 3 MALDI-Imaging Synapt G2-Si of DCA and HA. Although UA and undergo CID for structural elucidation prior to detection DEOXYCHOLIC ACID 202.6 198.9 isomeric, the two bile acids, previously been detected in brain. METHODS with the TOFMS. The T-Wave mobility device was calibrated for CHOLIC ACID 204.1 200.8 TW could easily be differentiated using ion mobility and detected estimated rotationally averaged CCSN2 measurements using CHENODEOXYCHOLIC ACID 209.0 205.5 0 Time Mass Spectrometry DT where spotted on to the mouse brain tissue section. 2.00 2.50 3.00 3.50 4.00 4.50 5.00 drift tube obtained CCSN2 measurements of ions produced LITHOCHOLIC ACID 208.6 204.6 from polyalanine. MS: Vion IMS Q-ToF and Synapt G2-Si URSODEOXYCHOLIC ACID 208.3 205.5 Fig. 5 Gas-phase separation optimization of the bile acid iso- Mode: ESI and MALDI (-VE) GLYCODEOXYCHOLIC ACID 199.8 196.9 Gas-Phase separation re-optimisation of bile acid isomers; Modeling and Theoretical CCS determination mers obtained in CO2 upon the Synapt G2-Si Capillary voltage: 2kV GLYCOCHOLIC ACID 202.7 200.0 effect of mobility gas alteration from N2 to CO2 Cone: 40V HYODEOXYCHOLIC ACID 209.6 206.5  Obtain structure from PubChem Source temperature: 110°C - TAUROCHENODEOXYCHOLIC ACID 208.4 204.9  Remove hydrogen to create deprotonated molecules [M-H] 100 Scan rate: 1 spectrum/s of the bile acids GLYCOCHENODEOXYCHOLIC ACID 201.9 198.0 Modeling and Theoretical CCS generation

 Run Gaussian Optimisation for starting structure and partial TAUROLITHOCHOLIC ACID 208.1 204.1 ESI DCA HA charges TAURODEOXYCHOLIC ACID 207.0 203.9

 Run Distance Geometry to generate a set of conformations. TAUROCHOLIC ACID 208.4 205.7 % The bile acids were infused at a concentration of 0.1ng/µL 8000 conformation limit was set for the deprotonated mole- (MeOH) and the signal attenuated with the DRE lens Vion G2-Si - cules of the bile acids Table 1 CCSN2 and CCSN2 of [M-H] of the Bile Acids investigated

 Run energy minimization for candidate low energy confor- MALDI-Imaging 0 mations 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 2 3  Theoretical CCS were obtained using the Trajectory Method Fig. 7 Sample conformations for the bile acids obtained from Isomeric bile acid mixtures were spotted on a 30 µm mouse 1 4 in MOBCAL and from preliminary N parameters using the 100 computational modeling for a) Glycodeoxycholic Acid, b) brain section mounted on a glass slide. The slide was spray 2 100 Projection Superposition Approximation method Taurodeoxycholic Acid, c) Deoxycholic Acid, and d) Chenode- coated with matrix using the SunChrom SunCollect Sprayer. oxycholic Acid. Bile acids that have more interactions between 30 coats were applied at a flow rate of 20 µL/min. MALDI DCA CDA 1. DCA the carboxylic or sulfonate end group on the tail structure with images were processed using High Definition Imaging software CDA the hydroxyl groups on the fused ring system correspond to 100 2. CA % v1.3. Matrix: 9-aminoacridine (0.5 mg/mL in 4:1 EtOH:H O). 2 UA HA 3. CDA the bile acids that fall closer to the lower bound on the theoretical range.

Ion Mobility 4. HA % % 0 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 Mobility bath gas: N2 (Vion) N2 (G2-Si) CO2 (G2-Si only) CONCLUSIONS

Ion mobility cell: ~3.0mbar ~3.0mbar ~3.0mbar 100  Good correlation was achieved between the two T- IMS Wave velocity: 850 m/s 900 m/s 900 m/s 0 Time 9.00 10.00 Wave ion mobility instruments used in this study. Trap Wave Height: 40-60V 40V 40V DCA  In combination with accurate mass measurement,

% UA the additional molecular descriptor of CCS can aid bile 0 Time acid ion identification Workflow 1.00 1.50 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 7.00 7.50 8.00 Fig. 6 Experimental CCS vs. theoretical CCS ranges for CCS  The results indicate that the addition of CCS ESI-MS was used to measure ion drift-times upon a hybrid ion Fig. 2 Overlaid drift times (ms) of DCA, CA, CDA and HA. UA has a values obtained in N2 drift gas. Theoretical values were ob- measurements to searchable databases within a 0 Time tained using the Trajectory Method (green) in MOBCAL and mobility/ quadrupole / oa-ToF MS (Vion IMS Q-ToF), ESI and very similar drift time to CDA and is different by only 1 scan (the 2.00 2.50 3.00 3.50 4.00 4.50 5.00 5.50 6.00 6.50 ‘steroidomic-type’ workflow increases the specificity MALDI-imaging was also used upon a hybrid quadrupole / ion flight time for the pusher frequency). The inset shows an overlay of from the PSA Method (blue). Theoretical conformations were and selectivity of bile acid analysis, improving the generated with distance geometry. mobility / oa-ToF MS (Synapt G2-Si). N was used as the the individual drift times (ms) for UA, CDA and HA using a higher T- confidence in identification compared to traditional 2 Structures of the bile acids. Wave ion mobility velocity of 1500 m/s. Fig. 4 Gas-phase separation optimisation for the bile acid iso- mobility gas in both instruments and CO2 in the Synapt G2-Si analytical approaches mers obtained in N2 upon the Synapt G2-Si. only.