A high-capacity LC/MS system for the bioanalysis of samples generated from plate-based metabolic screening.

HPLC/MS is a linear technique characterized by serial injection and analysis of individual samples. Parallel-format high-throughput screens for druglike properties present a significant analytical challenge. Analysis speed and system ruggedness are key requirements for bioanalysis of thousands of samples per day. The tasks involved in LC/MS analysis are readily divided into three areas, sample preparation/liquid handling, LC/MS method building/sample analysis, and data processing. Several automation and multitasking strategies were developed and implemented to minimize plating and liquid handling errors, reduce dead times within the analysis cycle, and allow for comprehensive review of data. Delivering multiple samples to multiple injectors allows the autosampler time to complete its wash cycles and aspirate the next set of samples while the previous set is being analyzed. A dual-column chromatography system provides column cycling and peak stacking and allows rapid throughput using conventional LC equipment. Collecting all data for a compound into a single file greatly reduces the number of data files collected, increases the speed of data collection, allows rugged and complete review of all data, and provides facile data management. The described systems have analyzed over 40 000 samples per month for two years and have the capacity for over 2000 samples per instrument per day.

AutoScan: an automated workstation for rapid determination of mass and tandem mass spectrometry conditions for quantitative bioanalytical mass spectrometry.

An automated flow injection analysis (FIA) mass spectrometry system (AutoScan) was developed to allow rapid unattended determination of optimal conditions during mass (ms) and tandem mass spectrometry (ms/ms) on new chemical entities (NCEs) arranged in 96-well plates. The 96-well plate is placed on the deck of a modified Gilson Multiprobe autosampler for injection into a PE Sciex API 2000 triple quadrupole mass spectrometer. A customized software interface is used to create the necessary scan experiments by associating each 96-well plate of NCEs to be scanned with an index file containing data on the identity of each analyte and its expected molecular weight. Analytes are injected four at a time into a custom injection manifold and conventional mass spectra are acquired in both polarities (+/-) using an alternating positive/negative Q1 scan function. The software determines the optimal polarity and definitive precursor ion for all analytes and uses the results to build the injection sequence for product ion scanning. The samples are automatically re-injected under MS/MS conditions, and product ion scans that loop among different collision energies are collected for each analyte. The resulting data are processed automatically and the optimal MS/MS transitions for each analyte are selected. A color-coded graphical interface facilitates data review. Any unusual ion transitions or transposition errors made during plate preparation are noted and corrected. Complete MS and MS/MS conditions are obtained for 96 compounds in about one hour and the resulting data are available for download as sample control injection sequence files.

Assay of ezlopitant, a substance P receptor antagonist, and metabolites in biological matrices by gas chromatography with mass spectrometric detection: simultaneous analysis of a benzyl alcohol and alkene.

A method for the analysis of the substance P antagonist ezlopitant and two active metabolites in serum using solid-phase extraction followed by GC-MS analysis is described. The linear dynamic range was 1.0 to 100 ng/ml and precision and accuracy over this range were within 15%. Upon injection of reconstituted sample extracts into the hot injector port of the gas chromatograph, the benzyl alcohol metabolite undergoes a small amount of spontaneous dehydration to the alkene metabolite. We have incorporated an additional hexadeuterated internal standard of the benzyl alcohol into the assay to permit measurement of the extent of dehydration in each sample. This novel approach should be generally applicable to the simultaneous determination of benzyl alcohols and corresponding alkenes by GC-MS when the possibility exists that the alcohol can undergo spontaneous dehydration to the alkene in the injector port of GC instrumentation.

High-throughput method development approaches for bioanalytical mass spectrometry.

A rational approach to the development and optimization of solid-phase extraction (SPE) methods is described. The semiautomated scheme allows for the simultaneous testing of multiple chemistries using a custom multiple-sorbent 96-well method development plate. Optimized extraction conditions for up to five analytes are determined in a single 2.5-h experiment. The experiment can be tailored to determine SPE conditions (including wash protocols) for related analytes. Data obtained by liquid chromatography-atmospheric pressure ionization-mass spectrometry allows the quantitation of absolute recovery and selection of the best extraction conditions for approximately 100 analytes of diverse structure. Optimized extraction protocols yielding at least 80% recovery are determined for 81% of the analytes. For 96% of the analytes screened, extraction conditions resulting in recoveries of > or = 60% are determined. The most generic set of SPE conditions consist of either C8 or C18 sorbent with an eluent composition of acetonitrile with 5mM nitric acid added.

Development and validation of a high-sensitivity assay for an antipsychotic agent, CP-88,059, with solid-phase extraction and narrow-bore high-performance liquid chromatography.

An analytical method has been developed and validated for the quantitation of CP-88,059 in human serum. The compound and internal standard were extracted from serum by solid-phase extraction with a weak cation-exchange phase. The analytes were resolved from endogenous interferences using narrow-bore (2.1 mm I.D.) C18 reversed-phase HPLC. Column effluent was monitored by UV absorbance detection at 215 nm. The standard curve range was 1 to 250 ng/ml. The accuracy and precision values for the method were within +/- 10% and +/- 15%, respectively. A four-fold detectability enhancement was achieved using a 2.1 mm I.D. HPLC column relative to the more common 4.6 mm I.D. column. A performance comparison was made between the 2.1 mm I.D. column used for validation and a 4.6 mm I.D. column with the same stationary phase.

High-performance liquid chromatographic determination of 2',3'-didehydro-3'-deoxythymidine, a new anti-human immunodeficiency virus agent, in human plasma and urine.

Sensitive and selective high-performance liquid chromatographic techniques have been developed for the determination of 2'-3'-didehydro-3'-deoxythymidine, d4T (BMY-27857), in human plasma and urine. The methods had linear standard curves over the concentration ranges 0.025-25.0 and 0.5-100 micrograms/ml for the plasma and urine matrices, respectively. Both methods used solid-phase extraction for isolating d4T and the internal standard, thymidine oxetane, from the biological matrix. In addition, the analytical column, mobile phase, instrumentation and chromatographic conditions used for both methods were identical. The ultraviolet absorbance of the column effluent was monitored at 266 nm. Results of analysis of quality control samples indicated that the intra-assay precision values, as measured by percent relative standard deviation, were within 12 and 3%, and accuracy samples deviated less than 10 and 5% from nominal values for the plasma and urine assays, respectively.

High-performance liquid chromatographic assay of alpha-(4-fluorophenyl)-4-(5-fluoro-2-pyrimidinyl)-1-piperazine butanol (BMY 14802), a potential antipsychotic drug, in monkey and rat plasma.

A high-performance liquid chromatographic-fluorescence method was developed for the quantitative analysis of BMY-14802 (I) in monkey and rat plasma. After the addition of the internal standard (BMY-14853 I.S.), 250 microliters of plasma were made basic by the addition of 2 ml of saturated sodium carbonate buffer. Compound I and the I.S. were then extracted into 5 ml of methyl tert.-butyl ether. The organic phase was evaporated and the resulting residue was reconstituted in mobile phase. Final separation and quantitation of I was achieved on an octadecyl column with a 0.05 M potassium phosphate-acetonitrile-triethylamine-85% phosphoric acid (650:350:0.1:0.05, v/v) mobile phase. Fluorescence detection was used to monitor the eluent at an excitation wavelength of 240 nm and an emission wavelength of 400 nm. The limit of detection was 0.5 ng/ml. The standard curve was linear over the range 5.0-1000 ng/ml. Intra-assay and inter-assay precision values were less than 4.0% relative standard deviation and accuracy was within 12% of nominal values. Compound I was shown to be stable in monkey and rat plasma for at least six months when stored at -20 degrees C.