Liquid Microjunction Surface Sampling Coupled with High-Pressure

Jun 18, 2010 - Additional information as noted in text. ... Spatially-Directed Protein Identification from Tissue Sections by Top-Down LC-MS/MS with E...
0 downloads 11 Views 1MB Size
Download PDF
Anal. Chem. 2010, 82, 5917–5921

Letters to Analytical Chemistry Liquid Microjunction Surface Sampling Coupled with High-Pressure Liquid ChromatographyElectrospray Ionization-Mass Spectrometry for Analysis of Drugs and Metabolites in Whole-Body Thin Tissue Sections Vilmos Kertesz* and Gary J. Van Berkel* Organic and Biological Mass Spectrometry Group, Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6131 In this work, a commercially available autosampler was adapted to perform direct liquid microjunction (LMJ) surface sampling followed by a high-pressure liquid chromatography (HPLC) separation of the extract components and detection with electrospray ionization mass spectrometry (ESI-MS). To illustrate the utility of coupling a separation with this direct liquid extraction based surface sampling approach, four different organs (brain, lung, kidney, and liver) from whole-body thin tissue sections of propranolol dosed and control mice were examined. The parent drug was observed in the chromatograms of the surface sampling extracts from all the organs of the dosed mouse examined. In addition, two isomeric phase II metabolites of propranolol (an aliphatic and an aromatic hydroxypropranolol glucuronide) were observed in the chromatograms of the extracts from lung, kidney, and liver. Confirming the presence of one or the other or both of these glucuronides in the extract from the various organs was not possible without the separation. These drug and metabolite data obtained using the LMJ surface sampling/HPLC-MS method and the results achieved by analyzing similar samples by conventional extraction of the tissues and subsequent HPLC-MS analysis were consistent. The ability to directly and efficiently sample from thin tissue sections via a liquid extraction and then perform a subsequent liquid phase separation increases the utility of this liquid extraction surface sampling approach.

drugs.1,2 Inherently, WBA cannot distinguish between the parent drug and the metabolites of that drug. Currently, the particular molecular forms and quantity of the drug-related material present must be determined from punched samples for areas of interest in the same tissue sections (e.g., radioactive “hot spots”) or from whole organ tissue homogenates from separate animals, which are analyzed with conventional sample extraction, cleanup and high-performance liquid chromatography-mass spectrometry (HPLC-MS) or tandem mass spectrometry (MS/MS). While this conventional procedure might be fully automated, a direct surface sampling, ionization, and analysis method by mass spectrometry of such thin tissue sections, including the same tissues used for WBA study, would save time and other resources by shortening the sampling and extraction steps of a quick first pass look for drug and metabolite distributions. Currently, the most widely used technique for molecular identification of drug related materials directly from animal thin tissue sections is matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS).3-5 While this technique continues to mature, some lingering limitations are apparent. First, MALDIMS requires the careful coating of the tissue with a matrix compound prior to the MS analysis. Also second, most MALDIMS analyses report detection of parent drug and some phase I metabolites, but apparently only rarely are phase II metabolites reported.4 This may be because phase II metabolites, like glucuronides and glutathione conjugates, are relatively fragile and do not survive the laser desorption/ionization process intact. Newer atmospheric pressure surface sampling/ionization tech-

In the area of drug discovery, the distribution of total drug related compounds in thin tissue sections is often determined using whole-body autoradiography (WBA) employing radiolabeled

(1) Solon, E. G.; Balani, S. K.; Lee, F. W. Curr. Drug Metab. 2002, 3, 451– 462. (2) Food and Drug Administration. www.fda.gov, 1005; 21CFR312.23. (3) Khatib-Shahidi, S.; Andersson, M.; Herman, J. L.; Gillespie, T. A.; Caprioli, R. M. Anal. Chem. 2006, 78, 6448–6456. (4) Stoeckli, M.; Staab, D.; Schweitzer, A. Int. J. Mass Spectrom. 2007, 260, 195–202. (5) Walch, A.; Rauser, S.; Deininger, S.-O.; Ho¨fler, H. Histochem. Cell Biol. 2008, 130, 421–434.

* To whom correspondence should be addressed. Phone, 865-574-4878; fax, 865-576-8559; e-mail, [email protected] (V.K.). Phone, 865-574-1922; fax, 865576-8559; e-mail, [email protected] (G.J.V.B.). 10.1021/ac100954p  2010 American Chemical Society Published on Web 06/18/2010

Analytical Chemistry, Vol. 82, No. 14, July 15, 2010

5917

niques,6-8 like desorption electrospray ionization (DESI)-MS,9 have also been demonstrated for determining drug related materials in thin tissue sections.10 This technique does not require a matrix coating, but initial reports from the analysis of tissue from drug dosed animals indicate that overall sensitivity may need improvement and detection of glucuronides known to be present in the tissue have so far been unsuccessful.11 Direct liquid extraction based surface sampling probes6 employ an alternative atmospheric pressure surface sampling/ionization technology and have shown success in the analysis of both drugs and phase II metabolites from animal thin tissue sections.12-14 These probes reconstitute or extract an analyte from a surface by contacting that surface with a confined liquid stream, a liquid microjunction, or a solvent droplet. The extract is subsequently analyzed by mass spectrometry using a conventional atmospheric pressure ionization source employing electrospray ionization (ESI) or atmospheric pressure chemical ionization (APCI). It may be the efficiency of the extraction and the larger surface area typically sampled (about a 0.5 mm to as much as 4 mm-diameter circle compared to an about a 200 µm-diameter laser spot typically used in MALDI) combined with the use of gentle, conventional electrospray ionization (ESI) that accounts for the success of this surface sampling/ionization approach for the detection of phase II metabolites. As an added benefit, these direct liquid extraction based surface sampling methods allow for the incorporation of a liquid based separation after the surface sampling process, a capability not possible with techniques like MALDI-MS or DESI-MS.6 This capability can be important when dealing with complex sample matrixes or targeted analyte components that cannot be easily distinguished on the basis of mass-to-charge ratio or by gasphase methods like tandem mass spectrometry. The use of a separation coupled with a sealing surface sampling probe (SSSP) for elimination of matrix components has been recently demonstrated by Spooner et al.15 in the analysis of drugs in dried blood spots on paper. In this letter, we show another important use for a post surface sampling separation process, namely, the separation of isomeric drug metabolites sampled from a thin tissue section. We have previously used a S-SSP or a liquid microjunction (LMJ)-SSP to sample and analyze thin tissue sections from propranolol dosed mice.12-14 In those studies, propranolol as well hydroxypropranolol glucuronide metabolite were detected. Because no separation was involved, whether the detected glucuronide metabolite signal was due to the presence of one or the other or both known hydrox(6) Van Berkel, G. J.; Pasilis, S. P.; Ovchinnikova, O. J. Mass Spectrom. 2008, 43, 1161–1180. (7) Harris, G. A.; Nyadong, L.; Fernandez, F. Analyst 2008, 133, 1297–1301. (8) Venter, A.; Nefliu, M.; Cooks, R. G. Trends Anal. Chem. 2008, 27, 284– 290. (9) Taka´ts, Z.; Wiseman, J. M.; Goldan, B.; Cooks, R. G. Science 2004, 306, 471–473. (10) Wiseman, J. M.; Ifa, D. R.; Zhu, Y. X.; Kissinger, C. B.; Manicke, N. E.; Kissinger, P. T.; Cooks, R. G. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 18120–18125. (11) Kertesz, V.; Van Berkel, G. J.; Vavrek, M.; Koeplinger, K. A.; Schneider, B. B.; Covey, T. R. Anal. Chem. 2008, 80, 5168–5177. (12) Van Berkel, G. J.; Kertesz, V.; Koeplinger, K. A.; Vavrek, M.; Kong, A. T. J. Mass Spectrom. 2008, 43, 500–508. (13) Van Berkel, G. J.; Kertesz, V. Anal. Chem. 2009, 81, 9146–9152. (14) Kertesz, V.; Van Berkel, G. J. J. Mass Spectrom. 2010, 45, 252–260. (15) Abu-Rabie, P.; Spooner, N. Anal. Chem. 2009, 81, 10275–10284.

5918

Analytical Chemistry, Vol. 82, No. 14, July 15, 2010

ypropranolol glucuronide metabolites11,16 could not be determined. For the present work, we adapted a commercially available autosampler to perform LMJ surface sampling from mouse wholebody thin tissue sections with subsequent HPLC separation and ESI tandem mass spectrometric analysis of the extract. The particular type of LMJ sampling we used might be referred to as a liquid droplet extraction mode which has been previously demonstrated for simple and relatively fast (30 s sample-to-sample) quantitative analysis of spotted sample arrays using our continuous flow LMJ-SSP.17 More recently, this mode was implemented on a commercially available Advion NanoMate chip-based infusion nanoESI system.14,18 In the present case, a simple syringe is used to dispense a droplet of extraction solvent onto the tissue surface, maintain contact with the droplet, and then subsequently aspirate the droplet and extracted material back into the syringe. This extract is then injected onto an HPLC column connected to an ESI source and triple quadrupole mass analyzer. For illustration, four different organs (brain, lung, kidney, and liver) from wholebody thin tissue sections of propranolol dosed and control mice were examined by this approach. In addition to observing the parent drug in the chromatograms from the extracts of all organs of the dosed mouse examined, two isomeric phase II metabolites of propranolol (an aliphatic and an aromatic hydroxypropranolol glucuronide) were detected with different retention times in lung, kidney, and liver. Comparisons are made between these data and results achieved by analyzing similar samples by conventional extraction of the tissues and subsequent HPLC-MS analysis. EXPERIMENTAL SECTION Chemicals. HPLC grade acetonitrile (ACN) and water were purchased from Burdick & Jackson (Muskegon, MI). Formic acid (FA) (g96% purity) was purchased from Sigma-Aldrich (St. Louis, MO). Propranolol hydrochloride (Acros Organics, Morris Plains, NJ) and 4-hydroxy propranolol glucuronide (TLC PharmaChem Inc., Mississauga, Ontario, Canada) were obtained commercially and used without further purification. Thin Tissue Section Preparation. Preparation and handling of mouse whole-body thin tissue sections on tape (propranolol, compound 1, administered intravenously via the tail vein at 7.5 mg/kg as an aqueous solution in 0.9% NaCl and sacrificed at 60 min postdose) have been described in detail elsewhere.11 Coregistration of the optical images and the surface spots sampled was accomplished manually. Instrumentation. A thin tissue section on tape was first secured with clear tape onto a 3 in. × 4 in., 1.2 mm thick glass slide, and then the glass slide was secured with clear tape onto the top of a typical 96-well plate. The plate was then placed onto a plate holder which was secured onto an MS2000 x, y, z-axis robotic platform (Applied Scientific Instrumentation Inc., Eugene, OR), see Figure 1 and Figure S1 in the Supporting Information. An HTC PAL autosampler (LEAP Technologies Inc., Carrboro, NC) was used for liquid handling during surface sampling (Figure 2, see the Supporting Information for details). The x, y, z stage and the autosampler were both secured onto the same 24 in. × (16) Salomonsson, M. L.; Bondesson, U.; Hedeland, M. J. Mass Spectrom. 2009, 44, 742–754. (17) Van Berkel, G. J.; Kertesz, V.; King, R. C. Anal. Chem. 2009, 81, 7096– 7101. (18) http://www.advion.com/ (accessed April 2, 2010).

Scheme 1. Structure, Mass-to-Charge Ratio, and Origin of Major Product Ions for Propranolol (1) and Metabolites (2a and 2b)

Figure 1. Photograph of the injection needle assembly of the HTCPAL autosampler showing the syringe needle in the surface sampling position above a thin tissue section on tape secured onto a glass slide/96-well plate for surface analysis.

Figure 2. Sequential steps of the surface sampling process showing (a) the syringe needle approaching the surface, (b) a liquid microjunction created between the needle and a liver thin tissue section (inset shows a magnified view), and (c) dissolved sample aspirated back into the needle and ready to be injected onto an HPLC column.

24 in. aluminum breadboard (Thorlabs, Newton, NJ). This was done to fix the position of the two x, y, z robots (i.e., the stage and the autosampler) to each other and to minimize shaking of the system during movement of the autosampler arm. Also, the vial holder of the autosampler was removed and replaced with an

in-house made “needle guide stop” (Figure 1) (see details in the Supporting Information). The autosampler was coupled to an Agilent 1200 highperformance liquid chromatography (HPLC) system (Agilent Technologies, Santa Clara, CA) and to a 4000 QTRAP mass spectrometer (AB SCIEX, Concord, Ontario, Canada) (see Figure S1 in the Supporting Information). Compounds extracted from the surface using 90/10/0.1 (v/v/v) water/ACN/FA were injected onto a Synergi Hydro-RP HPLC column (50 mm × 2 mm, 4 µm particle size; Phenomenex, Torrance, CA). HPLC separation solvents A and B were water and ACN, respectively, both with 0.1% (v/v) FA. The 5 min-long gradient separation included the following steps: 0-0.5 min, 90% A; 0.5-3 min, linear gradient to 35% A; 3-3.5 min, linear gradient to 10% A; 3.5-3.6 min, linear gradient to 90% A; and 3.6-5 min, 90% A. The solution flow rate was 200 µL/min. On the basis of our previous results using other liquid extraction based surface sampling systems13,14 and HPLC-MS11 to analyze similar tissues, three selected reaction monitoring (SRM) transitions were monitored using positive ion mode ESI with an emitter voltage of 5 kV and turbo sprayer heater temperature of 300 °C. These SRM transitions included m/z 260.1 f 183.1 (collision energy (CE) ) 27 eV; declustering potential (DP) ) 60 V) for compound 1, and m/z 452.1 f 116.1 and m/z 452.1 f 276.1 (both CE ) 35 eV and DP ) 60 V) for the hydroxypropranolol glucuronides (compounds 2a and 2b). Scheme 1 shows the compound structures and the monitored precursor and product ions. The dwell time was 50 ms for each transition monitored. RESULTS AND DISCUSSION A whole-body thin tissue section from a mouse that had been administered propranolol was examined. The specific areas sampled are annotated in the photograph of the tissue section shown in Figure 3a. The chromatograms obtained monitoring the specific SRM transitions for propranolol (compound 1, m/z 260.1 f 183.1) and the potential hydroxypropranolol glucuronide metabolites (compounds 2a and 2b, m/z 452.1 f 276.1) from the surface sampling extracts of brain, lung, liver, and kidney are shown in Figure 3b-e, respectively. Propranolol was observed in the chromatograms of the liquid microjunction surface sampling extracts from all four organs examined, with the highest levels Analytical Chemistry, Vol. 82, No. 14, July 15, 2010

5919

Figure 3. (a) Photograph of a propranolol dosed mouse (7.5 mg/ kg, I.V. dosed, sacrificed 1 h after dose) whole-body thin tissue section on tape showing sampled locations of (1) brain, (2) lung, (3) liver, and (4) kidney tissues. The surface spots in part (a) were analyzed using the autosampler/HPLC-MS surface sampling system. Signal levels for (black line) propranolol (1, m/z 260 f 183) and (gray line, 2a and 2b) hydroxypropranolol glucuronide (m/z 452 f 276) were recorded during a 5 min HPLC-MS run of (b) brain, (c) lung, (d) liver, and (e) kidney samples. Measured signal intensities were normalized individually for the drug and for the metabolite using the highest signal of a transition observed during the entire analysis of the four samples (3400 and 1600 cps, respectively).

recorded in the brain and lung. These results were in line with those we had observed previously using a S-SSP and a LMJ-SSP to analyze tissues of mice dosed following the same protocol as described here.11,13,14 Most importantly, the chromatograms revealed in all organs, except the brain, the presence of two different hydroxypropranolol glucuronide metabolites. These two glucuronides appeared at retention times of 2.86 and 3.32 min, respectively. In our prior surface sampling work, we detected hydroxypropranolol glucuronide in these same tissues but could not confirm if one or more particular isomers were present. Results of separate HPLC-MS analyses of homogenates of the four organs of interest confirmed the existence of these same two metabolites in lung, liver, and kidney (see experimental details and Figure S2 in the Supporting Information) at nearly the same relative abundances as our surface sampling results. Thus, the direct surface sampling results are consistent in both the identification and relative quantitation provided by the more timeconsuming and costly conventional tissue homogenation, extraction, and HPLC-MS analysis. Sampling the same organs of a 5920

Analytical Chemistry, Vol. 82, No. 14, July 15, 2010

control tissue section did not produce signal above background levels for either the parent drug or the metabolites (data not shown). We have assigned the metabolites at retention times of 2.86 and 3.32 min as compounds 2a and 2b, the aromatic and the aliphatic 4′-hydroxypropranolol glucuronides, respectively. These assignments were based on the following two observations. First, the retention time of commercially available 2a, 4′-hydroxypropranolol glucuronide, was measured as 2.86 min (see Figure S3 in the Supporting Information). Second, glucuronidation of 4′hydroxypropranolol examined by Salomonsson et al.16 was found to result in two isomeric forms of 4′-hydroxypropranolol glucuronide with different retention times when 4′-hydroxypropranolol was incubated with uridine 5′-diphosphoglucuronic acid (UDPGA, glucuronidation agent) and uridine 5′-diphosphoglucuronyl transferase (UDPGT, microsomal glucuronidation enzyme). Using separation conditions similar to those used here, they identified the metabolite with a shorter retention time as an aromatic O-glucuronide (glucuronic acid attached via the aromatic OH group, i.e., compound 2a), and the one with longer retention time as an aliphatic O-glucuronide (glucuronic acid attached via the aliphatic OH group, i.e., compound 2b). In addition, we noted that the product ion at m/z 116 (detected by SRM transition m/z 452.1 f 116.1) was more abundant for metabolite 2a than 2b (as can be observed in the chromatograms shown in Figure S4 in the Supporting Information). The product ion spectra of these two different glucuronides published by Salomonsson et al.16 also showed this same difference. CONCLUSIONS In this letter, we described the use of an autosampler/ HPLC-MS system for direct liquid microjunction surface sampling of thin tissue sections, via a liquid droplet extraction mode, with subsequent separation and mass spectrometric analysis of the extract. This sampling mode was demonstrated using the current software elements of the autosampler system with only minor modifications to the hardware. The diameter of the area sampled was approximately equivalent to the 1 mm diameter of the autosampler syringe needle used to dispense and retrieve a droplet of extraction solvent from the thin tissue sections examined. The system was used to sample four major organs of whole-body mouse thin tissue sections. Propranolol was detected in all organs, and two isomeric hydroxypropranolol glucuronide metabolites, at different retention times, were detected in lung, kidney, and liver of the same tissue. The metabolite with shorter retention time was identified as an aromatic O-glucuronide and the one with longer retention time as an aliphatic O-glucuronide. One can anticipate that various modifications, both to hardware and software, would improve the convenient use and analytical performance of this type of liquid extraction surface sampling system. Of immediate practical importance, however, might be the investigation of alternative chromatographic separation phases for the present application. In this work, a highly aqueous extraction solvent (90/10/0.1 (v/v/v) water/ACN/FA) was chosen to be compatible with the separation employing the Synergi HydroRP HPLC column which used a highly aqueous initial solvent composition for the gradient separation. Using what we know to be a more optimum, albeit highly organic, extraction solvent for the drug and metabolites of interest (20/80/0.1 (v/v/v)

water/ACN/FA)11,13,14 resulted in a significant fraction of the target metabolite passing through the column unretained. Therefore, in order to use such a highly organic extraction solvent, resulting in a higher concentration of the targeted species in the extract, one might consider use of a hydrophilic interaction liquid chromatography (HILIC) column (or preconcentration column) that typically utilizes a high organic mobile phase early in the separation gradient.19 Beyond the present application, other types of chromatography, for example, those aimed at separation of the enantiomers of chiral drugs and their metabolites,20,21 might be exploited in direct surface sampling HPLC-MS analysis of thin tissue sections. In general, this ability to directly and efficiently sample from thin tissue sections via a liquid extraction and then perform a subsequent liquid phase separation is expected to increase the utility of direct liquid extraction surface sampling as a tool for drug discovery purposes.

ACKNOWLEDGMENT Dr. Marissa Vavrek (Merck Research Laboratories, West Point, PA) is thanked for providing the whole-body mouse thin tissue sections and the HPLC-MS data of tissue homogenates. Dr. James Bradshaw (Oak Ridge National Laboratory, Oak Ridge, TN) is thanked for manufacturing the needle guide stop. Reed Hokanson (LEAP Technologies Inc., Carrboro, NC) and Dr. Richard King (Pharmacadence Analytical Services, LLC, Hatfield, PA) are thanked for initial assistance with the operation and software of the HTC PAL autosampler. The study was supported by the Battelle Memorial Institute Technology Maturation Fund. ORNL is managed by UT-Battelle, LLC for the U.S. Department of Energy under Contract DE-AC05-00OR22725.

(19) http://www.phenomenex.com/cms400min/hilic.aspx (accessed June 10, 2010). (20) Erny, G. L.; Cifuentes, A. J. Pharm. Biomed. Anal. 2006, 40, 509–515. (21) Perez, S.; Barcelo, D. Trends Anal. Chem. 2008, 27, 836–846.

Received for review April 12, 2010. Accepted June 14, 2010.

SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

AC100954P

Analytical Chemistry, Vol. 82, No. 14, July 15, 2010

5921