Anal. Chem. 2001, 73, 708-714
Fully Automated 96-Well Liquid-Liquid Extraction for Analysis of Biological Samples by Liquid Chromatography with Tandem Mass Spectrometry Sean X. Peng,* Todd M. Branch, and Salane L. King
Health Care Research Center, Procter & Gamble Pharmaceuticals, 8700 Mason-Montgomery Road, Mason, Ohio 45040
A fully automated high-throughput liquid-liquid extraction (LLE) methodology has been developed for preparation of biological samples using a 96-well LLE plate and a 96-channel robotic liquid handling workstation. The 96well LLE plate is made of a 96-well filter plate filled with inert diatomaceous earth particles, allowing continuous and efficient extraction of analytes between the aqueous biological sample and the organic extraction solvent. Two carboxylic acid-based protease inhibitor compounds with high and low levels of plasma protein binding were chosen for the development and application of the automated methodology. The LLE extracts of the plasma samples of the two compounds were analyzed by high-performance liquid chromatography with electrospray (ESI) tandem mass spectrometry (LC-MS/MS). The LC-MS/MS method was developed using a rapid gradient LC separation, followed by sample introduction through an ionspray interface in the negative ion mode and tandem mass spectrometric detection with selected reaction monitoring. In the optimized LLE method, a formate buffer solution was first loaded into a 96-well filter plate packed with inert diatomaceous earth material. Then crude plasma samples and a water-immiscible organic solvent, methyl ethyl ketone, were sequentially added to the LLE plate so that LLE would occur in the interface between the two liquid phases on the surface of individual particles in each well. The organic eluate containing extracted analytes was evaporated and reconstituted for LC-MS/MS analysis. This fully automated LLE methodology avoids several disjointed steps involved in a manual or semiautomated LLE method, leading to significantly reduced sample preparation time, increased sample throughput, and clean sample extracts for improved ESI-MS/MS detection. The automated LLE methodology is universal and can be employed for sample preparation of other biological fluids. The complete bioanalytical method, based on the automated LLE and fast gradient LC-MS/MS, was validated and successfully applied to the quantitative analysis of protease inhibitors in rat plasma. Recent advances in combinatorial chemistry and high-throughput screening, combined with today’s rapid developments in * Corresponding author: (phone) (513) 622-3944; (fax) (513) 622-3681; (e-mail)
[email protected].
708 Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
genomics, proteomics, and bioinformatics, have made a profound impact on all aspects of the drug discovery and development process. Many areas in drug discovery such as in vitro absorption and metabolism as well as pharmacokinetics and toxicology have adopted high-throughput methodologies to increase sample throughput and speed of respective screening assays. Similarly, all these have also revolutionized analytical approaches to supporting various areas of research programs. Improving sample throughput and automation has continued to play a key role in analytical sciences in the pharmaceutical industry in order to provide timely and quality analytical data to support drug discovery. Fast, accurate, and precise determination of drug compounds and related substances in complex biological matrixes is crucial in drug preclinical and clinical studies. In the bioanalytical arena, high-performance liquid chromatography with tandem mass spectrometry (LC-MS/MS) has been widely employed for the analysis of drug compounds in biological fluids because of its excellent specificity, speed, and sensitivity. However, the electrospray ionization process in a typical LC-MS/MS method is very sensitive to the interferences from the potential coelution of some endogenous components in the biological matrixes, affecting ionization efficiency, reproducibility, precision, and accuracy of the analysis.1-3 To minimize these interferences and improve overall ruggedness, an effective sample cleanup procedure should be utilized to make the best use of the LC-MS/MS technology. Sample preparation has been a critical step for accurate and reliable LC-MS/MS assays and has been the subject of several recent reviews.4-6 With a massive increase in the number of druglike compounds available from combinatorial chemistry and a potential explosion of the number of biological targets from genomics and proteomics, the preparation of drug compounds in biological fluids for bioanalysis can become a bottleneck in the process of drug discovery and development. Currently, the most widely employed biological sample preparation methodologies are solid-phase extraction (SPE), liquid-liquid extraction (LLE), and (1) Matuszewski, B. K.; Constanzer, M. L.; Chavez-Eng, C. M. Anal. Chem. 1998, 70, 882-889. (2) Buhrman, D. L.; Price, P. I.; Rudewicz, P. J. J. Am. Soc. Mass Spectrom. 1996, 7, 1099-1105. (3) Fu, I.; Woolf, E. J.; Matuszewski, B. K. J. Pharm. Biomed. Anal. 1998, 18, 347-357. (4) Henion, J.; Brewer, E.; Rule, G. Anal. Chem. 1998, 70, 650A-656A. (5) Rossi, D. T. LC-GC 1999, 17, S4-S8. (6) Hennion, M.-C. J. Chromatogr., A 1999, 856, 3-54. 10.1021/ac001036c CCC: $20.00
© 2001 American Chemical Society Published on Web 12/21/2000
protein precipitation (PPT). SPE has been a widely utilized technique for biological sample preparation, not only because of its ease for automation but also because of its ability to selectively remove interfering matrix components. Recent development and commercialization of 96-channel robotic liquid handling workstations7 as well as a wide selection of 96-well SPE sorbents8 afford the rapid development and automation of SPE methods to eliminate traditional time-consuming and labor-intensive biological sample preparation steps for plasma9-11 and urine samples.12 The advantages of sample preparation by SPE include the removal of nonvolatile salts and obtainment of a relatively clean extract with reduced interfering matrix components, allowing rapid LC-MS/ MS runs and improved sensitivity of MS detection. SPE automation in a 96-well format significantly reduces sample preparation time, increases sample throughput, and improves assay reproducibility. In contrast, PPT has been the main methodology for plasma sample preparation because of its simplicity and universality for all small drug molecules in plasma. PPT has been semiautomated using a robotic liquid handler and 96-well plates in several disjointed steps: precipitant addition, vortex mixing, centrifugation, and supernatant transfer.13 Recent development of fully automated PPT by filtration using 96-well filter plates and a robotic liquid handler has made the PPT method speedy and even more attractive.14,15 However, the PPT approach only removes plasma proteins, leaving other plasma matrix components in the sample. When an electrospray ionization (ESI) interface is employed for sample introduction and a rapid assay time is desired for higher throughput, coeluting matrix components can cause ion suppression and reduce the ion intensity of the analyte. This effect can also lead to decreased reproducibility and accuracy of the assay. LLE has been gaining popularity because it is highly selective and can provide very clean sample extracts, allowing reproducible and accurate MS detection.16-20 Compared to SPE and PPT, the main advantage of LLE for biological sample preparation is that it renders the least amount of ion suppression in ESI-MS/MS detection,21 leading to improved reproducibility and accuracy of the assay results. However, the lack of amenability to automation has limited its widespread use. Recent development and applica(7) Smith, G. A.; Lloyd, T. L. LC-GC 1998, 16, S22-31. (8) Wells, D. A. LC-GC 1999, 600-610. (9) Janiszewski, J.; Schneider, R. P.; Hoffmaster, K.; Swyden, M.; Wells, D.; Fouda, H. Rapid Commun. Mass Spectrom. 1997, 11, 1033-1037. (10) Davies, I. D.; Allanson, J. P.; Causon, R. C. J. Chromatogr., B 1999, 732, 173-184. (11) Peng, S. X.; King, S. L.; Bornes, D. M.; Foltz, D. J.; Baker, T. R.; Natchus, M. G. Anal. Chem. 2000, 72, 1913-1917. (12) Zhang, H.; Henion, J. Anal. Chem. 1999, 71, 3955-3964. (13) Watt, A. P.; Morrison, D.; Locker, K. L.; Evans, D. C. Anal. Chem. 2000, 72, 979-984. (14) King, S. L.; Foltz, D. J.; Baker, T. R.; Peng, S. X. Proceedings of 48th ASMS Conference on Mass Spectrometry and Allied Topics, Long Beach, CA, June 11-16, 2000. (15) Biddlecombe, R. A.; Pleasance, S. J. Chromatogr., B, 1999, 734, 257-265. (16) Ke, J.; Yancey, M.; Zhang, S.; Lowes, S.; Henion, J. J. Chromatogr., B 2000, 742, 369-380. (17) Jemal, M.; Teitz, D.; Ouyang Z.; Khan, S. J. Chromatogr., B 1999, 732, 501-508. (18) Ramos, L.; Bakhtiar, R.; Tse, F. L. S. Rapid Commun. Mass Spectrom. 2000, 14, 740-745. (19) Steinborner, S.; Henion, J. Anal. Chem. 1999, 71, 2340-2345. (20) Zhang, N.; Hoffman, K. L.; Li, W.; Rossi, D. T. J. Pharm. Biomed. Anal. 2000, 22, 131-138. (21) Bonfiglio, R.; King, R. C.; Olah, T. V.; Merkle, K. Rapid Commun. Mass Spectrom. 1999, 13, 1175-1185.
tion of a semiautomated 96-well LLE methodology16-20 have demonstrated that LLE is an attractive alternative to SPE in complex biological sample preparation. Further improvement in automation of LLE is apparently needed to take full advantage of LLE and make it comparable to SPE in speed and automation. In parallel to our recent development and application of automated 96-well LLE for the initial purification of combinatorial libraries,22 we here developed a fully automated 96-well LLE methodology for biological sample preparation. This fully automated LLE method avoided the use of three disjointed steps in the semiautomated LLE method: phase mixing by vortexing, phase separation by centrifugation, and phase transfer by aqueous layer freezing. Two structurally different matrix metalloprotease (MMP) inhibitor compounds that show high and low levels of plasma protein binding, respectively, were chosen here to illustrate the utility of the automated LLE methodology. Carboxylic acid-based MMP inhibitors are actively being explored as potential drug candidates for the treatment of cancers, arthritic disorders, and other connective tissue related diseases.23 In the screening and selection of these compounds for the targeted disease indications, the pharmacokinetic and toxicokinetic properties of each compound can play a key role in making a decision on whether to advance a compound for further studies. In support of pharmacology, toxicology, and animal efficacy studies, rapid, sensitive, and selective analytical methods are required for the routine determination of drug levels in biological fluids. Since a drug must have a sufficient amount available in the blood stream and be delivered to the site of action to produce its therapeutic effects, it is important to develop reliable bioanalytical methods to measure drug levels in plasma. This paper reports the development and application of a fully automated 96-well LLE methodology in conjunction with LC-MS/MS for the determination of two carboxilic acid-based MMP inhibitors in plasma. EXPERIMENTAL SECTION Materials. Carboxylic acid-based MMP inhibitors (shown in Figure 1), compound I, compound II, and an internal standard (IS), were synthesized by Procter & Gamble Pharmaceuticals (Mason, OH). The plasma protein binding of the two compounds and the internal standard was determined as described previously.24 HPLC-grade acetonitrile, methanol, formic acid, ammonium bicarbonate, ammonium formate, ammonium acetate, ammonium hydroxide, and mono- and dibasic sodium phosphate were purchased from J. T. Baker (Phillipsburg, NJ). Ethyl acetate, butyl acetate, methyl-tert-butyl ether, methyl ethyl ketone, and chloroform were obtained from Aldrich (Milwaukee, WI). Blank rat plasma was bought from Rockland (Gilbertsville, PA). A phosphate-buffered saline (PBS) solution at pH 7.4 was obtained from Life Technologies (Gaithersburg, MD). Inert diatomaceous earth material (Hydromatrix, high-purity and flux-calcined particles with an average diameter of 1 mm) and 96-well collection plates were purchased from Varian (Harbor City, CA) while 96-well hydrophobic GF/C glass fiber filter plates were obtained from Whatman (Clifton, NJ). (22) Peng, S. X.; Henson, C.; Strojnowski, M. J.; Golebiowski, A.; Klopfenstein, S. R. Anal. Chem. 2000, 72, 261-266. (23) Whittaker, M.; Floyd, C. D.; Brown, P. Gearing, A. J. H. Chem. Rev. 1999, 99, 2735-2776. (24) Peng, S. X.; Henson, C.; Wilson, L. J. J. Chromatogr., B 1999, 732, 31-37.
Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
709
Figure 1. Structures of I, II, and internal standard.
Figure 2. Schematic representation of the fully automated liquidliquid extraction.
Sample Preparation. Stock solutions of both the compounds and internal standard were prepared in methanol to give concentrations of 1.0 mg/mL and 100 µg/mL, respectively. An appropriate amount of each stock solution was evaporated to dryness under a stream of nitrogen. The dry residues were reconstituted in blank rat plasma to give a working solution of 1000 ng/mL for each compound or reconstituted in PBS buffer (pH 7.4) to yield a working solution of 40 ng/mL for the internal standard. Plasma standards were made by diluting the 1000 ng/mL working solution in blank plasma to yield 11 standard solutions ranging from 0.1 to 1000 ng/mL. Quality control (QC) samples were prepared as described for the calibration standards to give four concentration levels: 0.5 ng/mL (LLOQ QC), 5 ng/mL (low-level QC), 50 ng/ mL (mid-level QC), and 800 ng/mL (high-level QC). All above plasma standards and QC samples were subject to automated 96well LLE prior to being injected into the LC-MS/MS system. Automated 96-Well LLE Extraction Apparatus and Procedures. A 96-channel programmable liquid handling workstation (Quadra 96, model 320, Tomtec, Hamden, CT) and diatomaceous earth packed 96-well plates (∼260 mg/well) with hydrophobic GF/C glass fiber bottom filter were utilized for the automation of LLE for plasma samples (see Figure 2). The 96-well sample source plate was first prepared by transferring 100-µL aliquots of plasma samples including calibration standards and quality control samples to an empty plate. Then 25-µL aliquots of internal standard working solution were added to the 96-well sample source plate by the Quadra 96. A vacuum manifold system was used to house the LLE plate and collection plate. The LLE plate was placed on top of the vacuum manifold while a collection plate (collecting the eluate from the elution step) was put inside the manifold. The Quadra 96 was programmed to perform the LLE procedure as described below. The LLE plate was first loaded with 200-µL aliquots of an aqueous buffer solution by the Quadra 96. Then the Quadra 96 aspirated 100-µL aliquots of the sample from the 710
Analytical Chemistry, Vol. 73, No. 3, February 1, 2001
source plate to load onto the LLE plate. The LLE plate was then eluted with 800 µL of an organic extraction solvent to a 96-well collection plate filled with empty LC vials. During the elution step, very gentle vacuum (
