Applications of LC/M in Qualitative Drug Metabolite tudies
Case tudy: Identifying in vivo metabolites in human urine - sample cleanup and LC/M/M strategies Application detection of drug (Vanlev) metabolites in human urine limited sample need to identify major metabolites in high matrix background Approach on-line sample cleanup using small molecule trap column (maximize / ) Use instrument control language (ICL) on TQ to construct unique data-dependent acquisition modes using combined precursor, neutral loss, and product ion scan modes (maximize information per run)
Don t forget about sample cleanup!
Column witching ystem for Metabolite Cleanup ample Load Waste 2 Waters HPLC YMC-AQ C18 2.1x33mm 1 6 3 4 Gradient system 1 ml/min 5 HPLC Pump Wash solvent 0.1% TFA/H 2 Flow: 0.1 ml/min A Analytical Column YMC-AQ C18 4.6 x 150 mm Electrically Actuated Divert Valve Controlled by the Data ystem UV Finnigan TQ 700
LC/UV and Full-can LC/M Chromatograms of BM-186716 and its Metabolites in Human Urine Without Prior ample Clean-up 100 80 UV 220 nm 9.991 e+0 60 40 ormalized Intensity 20 0 100 80 RIC Full-can M 200-800 amu 1.345 e+07 60 40 20 0 8:20 16:40 25:00 33:20 Time (minutes)
LC/UV and Full-can LC/M Chromatograms of BM-186716 and its Metabolites in Human Urine Using an n-line Clean-up Column 100 80 60 UV 220 nm 9.991 e+0 ormalized Intensity 40 20 0 100 80 60 40 RIC Full-can M 200-800 amu 2.801 e+07 20 0 8:20 16:40 25:00 33:20 Time (minutes)
Extracted Ions of Metabolites of BM-186716 in Human Urine from Full-can LC/M Chromatograms: Before Use of Clean-up Column m/z= 423 6.693 e+04 m/z= 528 2.637 e+05 m/z= 585 1.272 e+04 m/z= 599 5.981 e+04 m/z= 615 3.856 e+04 m/z= 704 5.451 e+04 8:20 16:40 25:00 33:20 Time (minute s )
Extracted Ions of Metabolites of BM-186716 in Human Urine from Full- can LC/M Chromatograms: After Use of Clean-up Column m/z= 423 2.282 e+05 m/z= 528 6.401 e+05 m/z= 585 3.669 e+05 m/z= 599 1.958 e+05 m/z= 615 1.073 e+05 m/z= 704 1.867 e+05 8:20 16:40 25:00 33:20 Time (minutes )
Effect of Clean-up Column on bserved Molecular Ion pecies (Rt ~ 20min) Direct Injection on Analytical Column M+a + 621.1 M+H + M+K + 599.1 637.0 H 3 C H 6 H 8 C 6 2C Glucuronide Injection on Clean-up Column M+H + 599.1 M+a + 621.1 M+K + 637.0 500 550 600 650 700 m/z
Metabolite LC/M profiling in complex mixtures Traditional Approach Run LC/M and catalog potential metabolites (+16, +32, etc ). et up additional runs to collect product ion M/M data on all of these signals r, possibly collect both M and product ion M/M in a data-dependent mode in a single run Watch out: data-dependent product scans can be problematic in vivo metabolite samples due to endogenous material. A peak at every mass can lead to collection many useless product ion spectra which are not drug-related. Can perform precursor and neutral loss scans to search for relevant metabolites, but collection of data and sorting through signals is very labor intensive and time-consuming. If these are in vivo metabolites, you may run out of sample first!
Metabolite LC/M profiling in complex mixtures General precursor and neutral loss scanning approach for metabolite screening Carefully study fragmentation of parent drug Look for characteristic neutral losses or characteristic fragment ions in the product ion spectrum Use characteristic fragment ions to set up precursor scans Use characteristic neutral losses to set up neutral loss scans Add other neutral loss scans (e.g., L of 176 for glucuronidation) Run these LC/M/M experiments Perform product ion scans on any new potential metabolites with new LC/M/M experiments Can iterate through this approach as new metabolites are discovered Even more efficient! Use the precursor and neutral loss scans to trigger product ion scans Collect only potentially drug-related product ion scans Get more done with less sample in less time
Parent Ion/eutral Loss election H H 128 H 2 C BM-186716, DMPI eutral Loss of 176 gives glucuronide conjugates Parents of 128, other metabolites
Data-Dependent M/M Using Pre-Cursor Ion and eutral Loss canning to Detect Drug Metabolites LC/M/M Chromatogram 1 3 2 4 5 cans 1-2: Detect peak in neutral loss or pre-cursor ion scan mode cans 3-5: Collect product-ion scans Pre-set Threshold Time The TQ is scanned in either the neutral loss or pre-cursor ion scan modes The ICL procedure requires that the ion intensity of a detected mass surpass a pre-set threshold for two consecutive scans. When this condition is met, three product ion spectra for that mass are acquired.
BM-186716 and elected Metabolites (M+H) + = 585 H H 6 H 8 C 6 Glucuronide 2 C H 3 C H H 2 C (M+H) + = 423 CH 3 H 6 H 8 C 6 Glucuronide (M+H) + = 615 2 C H H H 2 C BM-186716 (M+H) + = 409 H 3 C H 6 H 8 C 6 Glucuronide 2 C (M+H) + = 599 H H 2 C H 2 H H H 2 C H 2 H H 2 C 6 H 8 C 6 Glucuronide 2 C (M+H) + = 528 (M+H) + = 704
Metabolite ID software can be used to assist with metabolite characterization
Metabolite ID input expected drug modifications
Metabolite ID Metabolite Data Browser - Extracted Modifications parent +16 +32 M M/M
Metabolite ID Metabolite Data Browser - M/M Correlation * *
High-Throughput LC/M Metabolic tability
High Throughput Metabolic tability Analyze many (thousands) of drug candidates to evaluate and rank compounds on the basis of metabolic stability Perform in vitro incubations w/ and wo/activator Use a fast LC/M system and method (1.5 min/sample) Export integrated peaks areas to spreadsheet and/or data base Ratio the signals from activated vs. non-activated in vitro metabolite incubations Visualize data using color coded sample status
High Throughput LC/UV/ELD/M Metabolic ystem tability ystem Pre-clinical Predictive Indicators Ion Trap M (Finnigan LCQ Duo) MAGIC 2002 Integrated HPLC, Finnigan LCQ Duo, Alcott A, integrated via Finnigan / custom software Integrated HPLC (MAGIC 2002) Autosampler (Alcott A) Flow: 1.5 mls/min Column: YMC D-AQ, 2X20mm, 5u M: Data directed, F and M/M UV: 200 and 220 nm, VWD 1.1 min method, 1.4 min full cycle, 750 samples/18 hr 0.05 min wide peaks 18-23 fully resolved peaks possible! Low flow rate, low delay volume, low psi nly slight decrease in theoretical peaks imilar cost to single quadrupole system Custom programming interface
Metabolic tability Buspirone - Human Liver Microsomes - Extracted Ions Parent Parent Internal tandard Internal tandard Activated - A on-activated -
Plate Level creening View ADME/TX - creen ummary - Metabolic tability Calculation PA - Parent-Activated P - Parent-on-Activated A - tandard-activated - tandard-on-activated 1 - (PA/A) (P/) Green = K Yellow =? Red = Bad
Conclusions In vivo metabolite samples still pose significant challenges for LC/M/M elucidation imple sample cleanup steps can result in dramatic improvement Intelligent data-dependent scanning can help to efficiently locate metabolites in the presence of matrix background High-throughput LC/M can be used to rank metabolic stability of compounds ew software tools (e.g., Metabolite ID) are useful in speeding the identification and characterization of drug metabolites
Acknowledgments Mark anders, Mike edved, Rich Gedamke (BM) Jeff Whitney (ovatia)