Application of a Liquid Extraction Based Sealing ... - ACS Publications

Application of a Liquid Extraction Based Sealing Surface Sampling Probe for Mass Spectrometric Analysis of Dried Blood Spots and Mouse Whole-Body Thin...
0 downloads 0 Views 282KB Size
Download PDF
Anal. Chem. 2009, 81, 9146–9152

Technical Notes Application of a Liquid Extraction Based Sealing Surface Sampling Probe for Mass Spectrometric Analysis of Dried Blood Spots and Mouse Whole-Body Thin Tissue Sections Gary J. Van Berkel* and Vilmos Kertesz* Organic and Biological Mass Spectrometry Group, Chemical Sciences Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831-6131 The utility of a liquid extraction based sealing surface sampling probe (SSSP) for the direct mass spectrometric analysis of targeted drugs and metabolites in dried blood spots (DBSs) and whole mouse thin tissue sections was demonstrated. The accuracy and precision for the quantitative analysis of a minimum of 50 ng/mL sitamaquine or acetaminophen in DBSs on paper were well within the required 15% dictated by internationally recognized acceptance criteria for assay validations. Analysis of wholebody mouse thin tissue sections from animals dosed with propranolol, adhered to an adhesive tape substrate, provided semiquantitative information for propranolol and its hydroxyproranolol glucuronide metabolite within specific organs of the tissue. The relative abundances recorded for the two compounds in the brain, lung, kidney, and liver were in nominal agreement with previously reported amounts based on analysis using a liquid microjunction surface sampling probe (LMJ-SSP), whole-body autoradiography (WBA), and high-pressure liquid chromatography-mass spectrometry (HPLC-MS). The ability to sample and analyze from tape-adhered tissue samples, which are generally employed in WBA analysis, presents the possibility of consecutive WBA and SSSP-MS analysis of the same tissue section. This would facilitate assignment, and possibly quantitation, of the different molecular forms of total drug related material detected in the WBA analysis. The flexibility to sample larger or smaller spot sizes, alternative probe sealing mechanisms, and a reduction in internal volumes and associated sample carryover issues will be among the first simple improvements necessary to make the SSSP-MS method a practical DBS and/or thin tissue section analysis toolortoexpanditsusetoothersurfacesamplingapplications. Direct liquid extraction based surface sampling probes1 are one segment of an emerging area in mass spectrometry (MS) * To whom correspondence should be addressed. Phone, 865-574-1922; fax, 865-576-8559; e-mail, [email protected] (G.J.V.B.). Phone, 865-574-4878; fax, 865-576-8559; e-mail, [email protected] (V.K.).

9146

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

that combines ambient surface sampling and ionization for analysis of analytes that are in or on surfaces.1-3 Direct liquid extraction surface sampling probes reconstitute or extract an analyte from a surface by contacting that surface with a confined liquid stream. That stream is both brought to the surface and is then carried on to the ionization source through a probe acting as a liquid conduit. In general, these types of probes might be coupled to any liquid introduction ionization source (e.g., electrospray ionization (ESI), atmospheric pressure chemical ionization (APCI), or another). Once in solution and carried into the source, analyte ionization is governed by the processes fundamental to the particular ionization source. Thus, these liquid extraction surface sampling probes can be applied to analysis situations in which the analyte can be dissolved and conducted into the probe and subsequently ionized by the respective ionization method being used. Currently, there are two basic types of liquid extraction surface sampling probes in use, namely, a “liquid microjunction” surface sampling probe (LMJ-SSP) used mainly by our group4 and a “sealing” surface sampling probe (SSSP) introduced by Luftmann.5 With the LMJ-SSP, the analyte is extracted from a surface by contacting that surface with a wall-less liquid microjunction between the sampling end of the probe and the surface. Analytical applications of the LMJ-SSP have involved the sampling and analysis of numerous compound types from diverse surfaces including stainless steel, glass, glassy carbon, and PEEK, hydrophobic paper, and reversed-phase thin layer chromatography (TLC) plates, as well as thin tissue sections on glass.4,6-15 (1) Van Berkel, G. J.; Pasilis, S. P.; Ovchinnikova, O. J. Mass Spectrom. 2008, 43, 1161–1180. (2) Harris, G. A.; Nyadong, L.; Fernandez, F. Analyst 2008, 133, 1297–1301. (3) Venter, A.; Nefliu, M.; Cooks, R. G. Trends Anal. Chem. 2008, 27, 284– 290. (4) Van Berkel, G. J.; Sanchez, A. D.; Quirke, J. M. E. Anal. Chem. 2002, 74, 6216–6223. (5) Luftmann, H. Anal. Bioanal. Chem. 2004, 378, 964–968. (6) Wachs, T.; Henion, J. Anal. Chem. 2001, 73, 632–638. (7) Wachs, T.; Henion, J. Anal. Chem. 2003, 75, 1769–1775. (8) Ford, M. J.; Van Berkel, G. J. Rapid Commun. Mass Spectrom. 2004, 18, 1303–1309. (9) Ford, M. J.; Kertesz, V.; Van Berkel, G. J. J. Mass Spectrom. 2005, 40, 866–875. 10.1021/ac901712b CCC: $40.75  2009 American Chemical Society Published on Web 10/09/2009

However, in the current form, this type of probe cannot be used to sample from porous surfaces (e.g., normal phase TLC plates or most papers) in which the liquid is conducted out from the probe into the wider surface rather than being aspirated back into the probe. With the Luftmann-type SSSP5 approach, analyte is extracted from a surface by sealing the probe to the surface using a knife edge on the probe that cuts into the surface. Note that this sealing mechanism does not readily allow analysis of analytes simply deposited on the surface of hard, inflexible, nonporous materials like metal, glass, or various plastics sheets. However, the Luftmann probe does allow the analysis of certain porous surfaces like alumina-backed normal phase TLC plates in its original format and analysis of glass-backed TLC plates after modifications to the original Luftmann sampler.16 Thus, the LMJSSP and SSSP approaches are in many ways complementary surface sampling techniques. To date, the Luftmann-type SSSP has been used for the analysis of relatively low molecular-mass compounds like caffeine or other pharmaceuticals or components of plant extracts separated on normal phase TLC plates.16-25 With the recent availability from CAMAG of a commercial version of the modified Luftmann-type SSSP named the “TLC-MS interface”,26 one can expect increased reporting of these same applications. However, as we will show here, this SSSP can also be applied to the analysis of other important analytical surfaces for which the knife edge sealing mechanism is viable. Among these surfaces are dried blood spots (DBSs) on paper and small animal whole-body thin tissue sections on adhesive tape. Dried blood spots are an accepted and convenient means to sample, process, store, and ship samples for analysis in newbornscreening programs.27,28 DBSs have also been used in humans for therapeutic drug monitoring and pharmacokinetic (PK) studies and are beginning to take hold as a routine sampling means for PK and toxicokinetic measurements in the drug discovery and clinical environments.29,30 In those studies, a small volume of blood (less than 100 µL) is deposited on paper and used for the analysis. (10) Ford, M. J.; Deibel, M. A.; Tomkins, B. A.; Van Berkel, G. J. Anal. Chem. 2005, 77, 4385–4389. (11) Asano, K. G.; Ford, M. J.; Tomkins, B. A.; Van Berkel, G. J. Rapid Commun. Mass Spectrom. 2005, 19, 2305–2312. (12) Kertesz, V.; Ford, M. J.; Van Berkel, G. J. Anal. Chem. 2005, 77, 7183– 7189. (13) Van Berkel, G. J.; Ford, M. J.; Doktycz, M. J.; Kennel, S. J. Rapid Commun. Mass Spectrom. 2006, 20, 1144–1152. (14) Van Berkel, G. J.; Kertesz, V.; Koeplinger, K. A.; Vavrek, M.; Kong, A. T. J. Mass Spectrom. 2008, 43, 500–508. (15) Kertesz, V.; Van Berkel, G. J.; Vavrek, M.; Koeplinger, K. A.; Schneider, B. B.; Covey, T. R. Anal. Chem. 2008, 80, 5168–5177. (16) Alpmann, A.; Morlock, G. Anal. Bioanal. Chem. 2006, 386, 1543–1551. (17) Aranda, M.; Morlock, G. Rapid Commun. Mass Spectrom. 2007, 21, 1297– 1303. (18) Luftmann, H.; Aranda, M.; Morlock, G. Rapid Commun. Mass Spectrom. 2007, 23, 3772–3776. (19) Morlock, G. E.; Jautz, U. J. Planar Chromatogr. 2008, 21, 367–371. (20) Dytkiewitz, E.; Morlock, G. E. J. AOAC Int. 2008, 91, 1237–1243. (21) Aranda, M.; Morlock, G. J. Chromatogr., A 2006, 1131, 253–260. (22) Morlock, G.; Ueda, Y. J. Chromatogr., A 2007, 1143, 243–251. (23) Aranda, M.; Morlock, G. J. Chromatogr. Sci 2007, 45, 251–255. (24) Jautz, U.; Morlock, G. J. Chromatogr., A 2006, 1128, 244–250. (25) Morlock, G.; Schwack, W. Anal. Bioanal. Chem 2006, 385, 586–595. (26) http://www.camag.de/ (accessed on July 14, 2009). (27) Chace, D. H. J. Mass Spectrom. 2009, 44, 163–170. (28) Chace, D. H.; Kalas, T. A.; Naylor, E. W. Clin. Chem. 2003, 49, 1797– 1817. (29) Spooner, N.; Lad, R.; Barfield, M. Anal. Chem. 2009, 81, 1557–1563.

Samples are punched, extracted, and cleaned up, and flow injection or high-pressure liquid chromatography-tandem mass spectrometry (HPLC-MS/MS) is used to quantify the targeted analytes. In the drug discovery process, whole-body autoradiography (WBA) using radiolabeled drugs is often applied for quantitative chemical imaging of total drug related compound in thin tissue sections.31,32 However, a major drawback of this technique is that it does not distinguish between the parent drug and a metabolite. Punched samples from these same sections, often secured on an adhesive tape backing, or whole organ tissue homogenates are used with conventional sample extraction, cleanup, and HPLC-MS or MS/MS to identify and quantify particular molecular forms of the drug-related material present. The rather involved analyses of both the DBSs and thin tissue section sections by mass spectrometry can be completely automated. However, the use of a surface sampling and ionization method to directly analyze the DBS or thin tissue section has the potential to simplify, speed up, and cut overall analysis costs. At least for thin tissue sections, some of these direct sampling/ ionization methods have already been attempted or demonstrated with varying degrees of success, including vacuum matrix assisted laser desorption ionization (MALDI)-MS,33,34 the ambient LMJSSP-MS,14 and desorption electrospray ionization (DESI)-MS15,35,36 techniques. In this technical note, we report for the first time the application of the commercially available Luftmann-type liquid extraction based SSSP to mass spectrometric analyses of drugs in DBS samples on paper, and drugs and metabolites in whole body thin tissue sections on adhesive tape. EXPERIMENTAL SECTION Chemicals. HPLC grade acetonitrile, methanol, and water were purchased from Burdick & Jackson (Muskegon, MI). Formic acid (g96% purity) was purchased from Sigma-Aldrich (St. Louis, MO). Propranolol hydrochloride was obtained from Acros Organics (Morris Plains, NJ). Sitamaquine [N,N-diethyl-N-(6-methoxy4-methylquinolin-8-yl)hexane-1,6-diamine] dihydrochloride and acetaminophen were obtained from GlaxoSmithKline (Greenford, U.K.) and Sigma-Aldrich, respectively. Isotopically labeled internal standards (ISs) sitamaquine-d10 dihydrochloride and acetaminophen-d4 were produced by Isotope Chemistry, GlaxoSmithKline (Stevenage, U.K.). Dried Blood Spot Sample Preparation. Stock solutions (1 mg/mL) of sitamaquine and sitamaquine-d10 were prepared in 50/50 (v/v) methanol/water. Rat blood was mixed with the analyte stock solutions prior to spotting to obtain the required sitamaquine (0-10000 ng/mL) and sitamaquine-d10 (570 ng/ mL) concentrations. Stock solutions (1 mg/mL) of acetamin(30) Edelbroek, P. M.; van der Heijden, J.; Stolk, L. M. L. Ther. Drug Monit. 2009, 31, 327–336. (31) Solon, E. G.; Balani, S. K.; Lee, F. W. Current Drug Metab. 2002, 3, 451– 462. (32) Food and Drug Administration. www.fda.gov, 1005; 21CFR312.23. (33) Khatib-Shahidi, S.; Andersson, M.; Herman, J. L.; Gillespie, T. A.; Caprioli, R. M. Anal. Chem. 2006, 78, 6448–6456. (34) Stoeckli, M.; Staab, D.; Schweitzer, A. Int. J. Mass Spectrom. 2007, 260, 195–202. (35) Wiseman, J. M.; Ifa, D. R.; Venter, A.; Cooks, R. G. Nat. Protoc. 2008, 3, 517–524. (36) Wiseman, J. M.; Ifa, D. R.; Zhu, Y.; Kissinger, C. B.; Manicke, N. E.; Kissinger, P. T.; Cooks, R. G. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 18120–18125.

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9147

Scheme 1. Structure and Mass-to-Charge Ratio of Analytes and Origin of Major Product Ions

ophen and acetaminophen-d4 were prepared in dimethyl formamide. Human blood was mixed with the analyte stock solutions prior to spotting to obtain the required acetaminophen (0-5000 ng/mL) and acetaminophen-d4 (7000 ng/mL) concentrations. Aliquots (15 µL) of the dosed rat (sitamaquine) or human (acetaminophen) blood samples were spotted onto Ahlstrom 237 filter paper (Ahlstrom Corp., Helsinki, Finland) for both the calibration standards and quality control (QC) samples and allowed to dry at room temperature for at least 2 h. One replicate of each member of a calibration standard set, which included a blank (no analyte, only IS) and total blank (no analyte, no IS) samples, was analyzed before and after six replicates of each member of the QC samples. Both the standards and QC samples were analyzed in order of decreasing analyte concentration. Thin Tissue Section Preparation. Male CD-1 mice (Charles River Laboratories) were administered propranolol intravenously via the tail vein at 7.5 mg/kg as an aqueous solution in 0.9% NaCl. At 60 min postdose, mice were euthanized with an isoflurane overdose and immediately frozen in dry ice/hexane. The frozen mice were embedded/blocked in 2% aqueous carboxymethyl

cellulose. Sagittal whole-body cryosections (40 µm thick) were prepared using a Leica CM3600 cryomacrotome and collected onto adhesive tape (8 in. × 3 in., 810 Scotch Brand Magic Tape, 3M) then freeze-dried within the chamber of the cryomacrotome. Prior to analysis using the SSSP, tissue sections were stored in a desiccator at room temperature. Color images of the tissue sections were acquired using a HP Scanjet 4370 flat-bed scanner (Hewlett-Packard, Palo Alto, CA). Details regarding the WBA analysis of these tissue sections have been presented elsewhere.15 Sealing Surface Sampling Probe Analysis. The inlet and outlet of the SSSP (TLC-MS interface, CAMAG, Muttenz, Switzerland) were coupled to an ACQUITY UltraPerformance Liquid Chromatography (UPLC) system (Waters Corporation, Manchester, U.K.) and to a 4000 QTRAP mass spectrometer (MDS SCIEX, Concord, Ontario, Canada), respectively (see Figure S1 in the Supporting Information). The gas pressure applied via a piston to the plunger of the SSSP interface was about 5 bar. Extraction solvent used in a particular experiment was selected to provide the maximum signal-to-noise ratio for that particular analyte. Tested solvent systems included water, acetonitrile, methanol, and various mixtures of these pure solvents, with and without 0.1% formic acid. The flow rate of the extraction solvent was 200 µL/min in all cases. For positive ion mode ESI, the emitter voltage was 5 kV and the turbo sprayer heater temperature was 300 °C. For positive ion mode APCI, the heated nebulizer probe temperature and needle current were 350 °C and 5 µA, respectively. Scheme 1 shows the compound structures and the monitored precursor and product ions. Ionization methods, extraction solvents, and MS conditions used for detection of each compound are listed in Table 1. The dwell time was 50 ms for each transition monitored. RESULTS AND DISCUSSION Operation of the SSSP. The extraction solvent flow paths and the individual steps of the surface sampling process with the SSSP are illustrated in Scheme 2. The solvent inlet capillary to the SSSP was connected to the UPLC pump, while the outlet capillary was connected to the ion source (ESI or APCI) of the mass spectrometer. At the beginning of a surface sampling experiment, a sample was positioned under the SSSP plunger while the extraction solvent flowed from the UPLC through a loop into the ion source (Scheme 2a). Next, the SSSP stainless steel plunger, with a 4 mm diameter disk shape cutting edge, was lowered to seal the sampling area by cutting into and compressing the sample substrate (paper or adhesive tape in this case) (Scheme

Table 1. Analytes, Ionization Methods, Extraction Solvent, Selected Reaction Monitoring Transitions, and Mass Spectrometer Parameters Settings Including Declustering Potential (DP) and Collision Energy (CE) analyte propranolol

extraction solvent 80/20/0.1 (v/v/v) acetonitrile/ water/formic acid

ionization method ESI+

hydroxypropranolol glucuronide sitamaquine

100/0.1 (v/v) methanol/ formic acid

APCI+

sitamaquine-d10 acetaminophen acetaminophen-d4

9148

methanol

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

ESI+

Q1 (m/z)

Q3 (m/z)

DP (V)

CE (eV)

260.1

183.1

60

27

452.1

276.1

60

35

344.4

271.2

42

30

354.4

271.2

42

30

152.1 156.1

110.1 114.1

24 24

40 40

Scheme 2. Individual Steps of the Surface Sampling Process with Expanded Views of the Probe Sealed at the Surfacea

a

Arrows and the black line show the direction of the liquid flow.

2b). With the valve switched from standby to extraction mode, the solvent previously being pumped directly to the ion source was diverted to the surface, where it extracted the analyte, and any other soluble components, and carried them into the ion source (Scheme 2c). After a suitable extraction time (typically 60 s), the valve was switched back to direct the extraction solvent from the UPLC through the loop into the ion source (Scheme 2d). With the gas pressure of 5 bar applied via a piston to the plunger, and the surfaces and solvents being used, we found this seal would hold a backpressure up to about 13 bar without leaking significantly. Leaking was not an issue unless the frit on the downstream side of the sampling probe became clogged. In the final step of the surface sampling process, the plunger was raised up from the surface (Scheme 2e) and then a cleaning station was positioned under the probe. Here, with the use of push button control, the sampling face of the plunger was cleaned of accumulated particulate material using a stream of pressurized nitrogen. In the case of the tissue sections on tape, the probe actually cut out a section from the sample which stayed with the probe until removed by this gas jet cleaning. Because of the probe design, complete elimination of sample carryover required extraction at a blank spot (typically 60 s) on the sampling surface (or another blank surface). This washed the SSSP plunger face and tubing from the plunger to the valve and on into the ion source. Given 15 or 30 s to position the probe for analysis of a DBS or tissue sample spot, respectively, a 60 s extraction time and 60 s wash time to eliminate carryover, spot-to-spot analysis time was either 2.5 or 3 min. Dried Blood Spot Analysis. Figure 1a shows the selected reaction monitoring (SRM) ion current chronogram for sitamaquine obtained from the analysis of each member of a calibration standard set of DBSs prepared from rat blood. With the use of positive ion mode ESI, each DBS sample was extracted for 60 s with a 100/0.1 (v/v) methanol/formic acid solvent at a flow rate of 200 µL/min. Inspection of these data showed the

Figure 1. SRM ion current chronograms of (a) sitamaquine and (b) sitamaquine-d10 obtained from the analysis of dried rat blood spot calibration standards using positive ion mode ESI. The concentration of sitamaquine is shown in part a. (c) Calibration curve constructed using calibration standards (line) and average of QC samples (O) with error bars (CV) using the ratio of background corrected integrated (over the 60 s sampling period) SRM signal of sitamaquine (10-10000 ng/mL) and that of sitamaquine-d10 (570 ng/mL) (Asitamaquine/Asitamaquine-d10) as a function of sitamaquine concentration (csitamaquine) in the blood spotted onto the paper substrate. The calibration data was analyzed using a least-squares regression with a 1/csitamaquine weighting and fit the model of Asitamaquine/Asitamaquine-d10 ) (1.63 × 10-3)csitamaquine - (1.12 × 10-2) (r2 ) 0.999). Statistical results obtained for the QC samples are summarized in Table 2.

analyte extraction profiles to have an asymmetry similar to that which might be expected from a flow injection experiment.37 Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9149

Table 2. Nominal (csitamaquine) and Calculated Mean (ccalcd,sitamaquine) Concentrations, Precision (% CV, n ) 6), and Accuracy (% Bias) of Sitamaquine Quality Control DBS Samples Based on a Linear Fit of Calibration Standards csitamaquine (ng/mL)

ccalcd,sitamaquine (ng/mL)

precision (% CV)

accuracy (% bias)

10000 5000 2000 1000 500 200 100 50 20 10

9851.9 5030.0 2159.4 1012.0 513.6 208.7 100.0 55.4 24.5 15.6

0.4 0.6 0.3 0.7 0.3 0.9 0.7 0.6 1.1 3.4

-1.5 0.6 8.0 1.2 2.7 4.4 0.0 10.8 22.6 55.7

However, at least for the higher level standards, the signal for the drug was truncated before it completely reached background levels. This truncation was caused by switching the solvent flow path after the 60 s sampling period. To clear out this remaining material from the probe, and to clean the sampling head, the system was washed for 60 s between each standard/sample spot by sampling a blank spot on the DBS sample card. One also observed in these chronograms a small peak after almost all individual spot samples resulting from this washout. The typical extraction profile and spot-to-spot signal reproducibility are more easily seen in and carryover is more easily estimated from Figure 1b, which presents the SRM signal for 11 replicates of the IS sitamaquine-d10 (570 ng/mL) collected simultaneously with the data shown in Figure 1a. The integrated area of the carryover peak observed when sampling a blank spot was approximately 3% of that of the SRM signal recorded when sampling a sample spot. The coefficient of variation (CV) calculated from the integrated area of these IS peaks was 3.6%. With the use of calibration standard data recorded before and after the analysis of QC samples, a calibration curve was constructed using the ratio of background corrected integrated SRM signal of sitamaquine and that of sitamaquine-d10 (Asitamaquine/Asitamaquine-d10) (over the 1 min sampling period) as a function of sitamaquine concentration (csitamaquine). The resulting calibration curve is plotted as the solid line in Figure 1c. The averaged data (O) and associated error bars (CV) obtained from the separate analysis of the QC samples (n ) 6) are also plotted in Figure 1c. As summarized in Table 2, the precision of measurements (CV) was less than about 1% down to the 20 ng/mL QC samples, increasing to 3.4% for the 10 ng/mL samples. The back calculated concentrations for the QC standards generally showed an accuracy within 10% down to the 50 ng/mL level. Thus, this method provides the required accuracy and precision within internationally recognized acceptance criteria for assay validations38 down to the 50 ng/mL level. Analysis of DBS samples from human blood containing acetaminophen was also accomplished. We found that best results (37) Pai, S. C.; Lai, Y. H.; Chiao, L. Y.; Yu, T. J. Chromatogr., A 2007, 1139, 109–120. (38) Shah, V. P.; Midha, K. K.; Findlay, J. W. A.; Hill, H. M.; Hulse, J. D.; McGilveray, I. J.; McKay, G.; Miller, K. J.; Patnaik, R. N.; Powell, M. L.; Tonelli, A.; Viswanathan, C. T.; Yacobi, A. Pharm. Res. 2000, 17, 1551– 1557.

9150

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

Figure 2. SRM ion current chronograms of (a) acetaminophen and (b) acetaminophen-d4 obtained from the analysis of dried human blood spot calibration standards using positive ion mode APCI. The concentration of acetaminophen is shown in part a. (c) Calibration curve constructed using calibration standards (line) and average of QC samples (O) with error bars (CV) using the ratio of background corrected integrated (over the 1 min sampling period) SRM signal of acetaminophen (50-5000 ng/mL) and that of acetaminophen-d4 (7000 ng/mL) (Aacetaminophen/Aacetaminophen-d4) as a function of acetaminophen concentration (cacetaminophen) in the blood spotted onto the paper substrate. The data was analyzed using a least-squares regression with a 1/cacetaminophen weighting and fit the model of Aacetaminophen/ Aacetaminophen-d4 ) (1.42 × 10-4)cacetaminophen + (7.33 × 10-3) (r2 ) 1.000). Statistical results obtained for the QC samples are summarized in Table 3.

were obtained for analysis of this compound using 100% methanol as the extraction solvent (200 µL/min) and positive ion mode APCI. Figure 2a shows the SRM ion current chronogram of acetaminophen for a calibration standard set. The analyte extraction profiles have the same basic appearance as those shown for sitamaquine in the experiments described above. Figure 2b presents the SRM signal for the IS acetaminophen-d4 (7000 ng/ mL) collected simultaneously with that of acetaminophen. The CV calculated from the integrated area of these IS peaks was 6.0%. A calibration curve constructed from the calibration standard data, recorded before and after the analysis of QC samples, using the ratio of background corrected integrated SRM signal of acetaminophen and that of its IS (Aacetaminophen/ Aacetaminophen-d4) (over the 1 min sampling period) as a function of acetaminophen concentration (cacetaminophen) is presented as the solid line in Figure 2c. The averaged data (O) and associated error bars (CV) obtained from the QC sample analysis (n ) 6) are also shown in Figure 2c. Statistical evaluation of the data summarized in Table 3 shows that the precision of measurements (CV) was less than about 2% down to the 100 ng/mL QC samples, increasing to 4.1% for the 50 ng/mL samples. The back calculated concentrations for the QC standards showed an accuracy within 4% down to the lowest investigated 50 ng/mL level. These quantitation metric results show that this method provides

Table 3. Nominal (cacetaminophen) and Calculated Mean (ccalcd,acetaminophen) Concentrations, Precision (% CV, n ) 6), and Accuracy (% Bias) of Acetaminophen Quality Control DBS Samples Based on a Linear Fit of Calibration Standards cacetaminophen (ng/mL)

ccalcd,acetaminophen (ng/mL)

precision (% CV)

accuracy (% bias)

5000 2000 500 200 100 50

4998.7 1984.8 510.7 206.4 99.7 51.4

0.4 0.9 1.0 1.3 2.0 4.1

0.0 -0.8 2.1 3.2 -0.3 2.7

accuracy and precision values required by internationally recognized acceptance criteria for assay validations down to the 50 ng/ mL acetaminophen level. Thin Tissue Section Analysis. Spot sampling experiments were used to demonstrate the detection of targeted exogeneous compounds in mouse whole-body thin tissue sections. For this investigation, we used tissues sections from a mouse that had been administered 7.5 mg/kg propranolol intravenously via the tail vein and sacrificed 60 min later. The tissue section was collected onto an adhesive tape substrate, which is a common format to collect whole-body tissue sections for WBA analyses.39 Using positive ion mode ESI and SRM detection, each sample spot was extracted for 1 min using a 80/20/0.1 (v/v/v) acetonitrile/water/formic acid at a flow rate of 200 µL/min. The areas sampled are annotated in the photograph of the tissue shown in Figure 3a. Note that the probe actually cut out a 4 mm diameter section from the tissue during the sampling process at each spot. These sample cutouts are not visible in the photograph because it was taken before the sampling process. The SRM chronograms obtained for propranolol and its hydroxypropranolol glucuronide metabolite from the sequential sampling of the different organs are shown in parts b and c of Figure 3, respectively. High propranolol signals were observed for brain, kidney, and lung (spots 2, 3, and 6, respectively). In addition, a significant level of the drug was detected in the stomach (spot 4). A much lower propranolol signal was recorded in the liver (spot 1) and lung (spot 5). As might be expected, signal levels for hydroxypropranolol glucuronide were highest in the liver and kidney (spots 1 and 3, respectively). Glucuronidation occurs mainly in the liver, making the drug more water-soluble to enhance subsequent elimination from the body by urination via the kidneys.40 Much lower hydroxypropranolol glucuronide signals were observed in the heart and lung (spots 5 and 6, respectively), and this metabolite was not detected in the brain (spot 2) or stomach (spot 4). These general observations are in line with the previously reported LMJ-SSP-MS/MS and WBA analyses of brain, liver, kidney, and lung tissues of a mouse dosed following the same protocol as described here.15 To show reproducibility of the sampling and signal levels from the various locations on the tissue, we examined another tissue section taken adjacent to and visually undistinguishable from the (39) Drug Discovery and Evaluation: Safety and Pharmacokinetic Assays; Vogel, H. G., Hock, F. J., Maas, J., Mayer, M., Eds.; Springer: Berlin, Germany, 2006; pp 587-590. (40) Sten, T.; Qvisen, S.; Uutela, P.; Luukkanen, L.; Kostiainen, R.; Finel, M. Drug Metab. Dispos. 2006, 34, 1488–1494.

Figure 3. (a) Photograph of a propranolol dosed mouse whole-body thin tissue section on adhesive tape. The six discrete points analyzed are annotated: 1 ) liver; 2 ) brain; 3 ) kidney; 4 ) stomach/contents; 5 ) heart; 6 ) lung. SRM ion current chronograms for (b) propranolol and (c) hydroxypropranolol glucuronide recorded during a 60 s sampling period at each point using positive ion mode ESI. Average integrated area with error bars (CV) of the SRM signals of (d) propranolol and (e) hydroxypropranolol glucuronide for organs analyzed in two separate tissue sections.

one examined above. SRM signals of the parent drug and that of the metabolite collected during the 1 min sampling period were integrated individually for all organs analyzed. Parts d and e of Figure 3 show the averaged integrated areas and associated error bars (CV) calculated from the analysis results of the two tissues sections for propranolol and hydroxypropranolol glucuronide, respectively. The precision of measurements was typically around 10% but always better than 20%. Though signal reproducibility is good, more work will be required to determine how well these signal levels reflect the actual quantities of the drugs and metabolites in each tissue. One might expect, for example, that the different composition of the various tissues might have a different matrix effect related to either extraction or ionization of these or any other targeted compounds. CONCLUSIONS In this technical note, we described the application of a liquid extraction based SSSP for mass spectrometric analyses of targeted drugs in DBS samples on paper, and targeted drugs and metabolites in thin tissue sections on an adhesive tape. Analysis Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

9151

of DBS samples employed both ESI and APCI. Quantitative evaluation of DBS samples containing a minimum of 50 ng/mL sitamaquine or acetaminophen resulted in accuracy and precision within internationally recognized acceptance criteria for assay validations. We also analyzed, using the SSSP and ESI, wholebody thin tissue sections of mice intravenously dosed with pharmacologically relevant levels of the drug propranolol. Analysis of the thin tissue sections adhered to an adhesive tape substrate provided at the minimum a semiquantitative abundance of propranolol and hydroxyproranolol glucuronide in the individual organs sampled. These relative abundance data were consistent with previous LMJ-SSP-MS/MS and WBA studies of tissue sections from mice subjected to the same drug administration protocol.15 The successful sampling of tape-adhered tissue samples, which are generally employed in WBA analysis, demonstrated the possibility of a consecutive WBA and SSSP-MS analysis. This would facilitate molecular characterization of the total drug related material distributions provided by WBA. Although the attempted analyses performed here with the SSSP were largely successful, one can anticipate that detrimental effects due to direct sampling of the surfaces, i.e., the lack of sample cleanup, may be problematic in other analysis situations. A simple example is the fact that the ion source and atmospheric pressure inlet of the mass spectrometer become more quickly contaminated when analyzing these types of samples than those which have gone through a cleanup procedure. When APCI is used, material associated with the DBSs was observed to rapidly build up on the discharge needle, which was positioned directly in the path of the heated nebulizer of this ionization source. The needle required cleaning after every complete set of standards and QC samples to maintain optimum performance (about 60-80 analyses). As another example, we analyzed sitamaquine and acetaminophen samples of a limited concentration range prepared on paper the same way as the DBS samples but without the blood matrix (data not shown). While we did not conduct a full statistical analysis on these data, we found the matrix-free sample signal levels to be approximately double those observed with the blood matrix present. The matrix suppression of the signal of other analytes might potentially prohibit a successful analysis at the required concentration of the analyte. The lack of a chromatographic separation of course may also inhibit differentiating certain isobaric species that might be present. An example would be propranolol, which has at least three possible isomeric glucuronides.41 With the current approach, these cannot be distinguished using simple SRM. Various simple changes to the SSSP used here would improve the analytical performance of the system. A smaller (and maybe

9152

Analytical Chemistry, Vol. 81, No. 21, November 1, 2009

even a larger) sampling probe plunger (currently 4 mm in diameter) would be of benefit. A smaller plunger specifically, say 1 mm in diameter, would allow sampling distinct features within individual organs of whole body tissue sections. Shorter, lower volume tubing connections between the sampling plunger and the switching valve would limit sample dilution and reduce washout time between samples possibly improving detection levels and sample-to-sample analysis speed. Additionally, a probe with a different sealing mechanism (e.g., an O-ring versus knife edge seal) could be used to sample from many other types of surfaces. This might include any number of hard, nonporous materials like glass, Teflon, or various plastics, or soft, sometimes porous, biological materials such as skin or other living tissue. Though with a significant degree of technical complexity added, it would be possible to couple the SSSP to an HPLC-MS system to integrate sampling and online separation to overcome sample matrix effects or isobaric overlaps. These simple advancements among others would improve the use of the SSSP-MS method for DBS and/or thin tissue section analysis and help broaden its application to the analysis of other analytically important surfaces. ACKNOWLEDGMENT Dr. Marissa Vavrek (Merck Research Laboratories, West Point, PA) is thanked for the whole-body mouse thin tissue sections. Dr. Neil Spooner and Dr. Paul Abu-Rabie (GlaxoSmithKline Research and Development, Ware, U.K.) are thanked for the dried blood spot samples. Dr. Eike Reich (CAMAG, Muttenz, Switzerland) is thanked for the loan of the TLC-MS interface. Study of the fundamentals of surface sampling using the sealing surface sampling probe employed here was supported by the Division of Chemical Sciences, Geosciences, and Biosciences, Office of Basic Energy Sciences, United States Department of Energy. ORNL is managed by UT-Battelle, LLC for the U.S. Department of Energy under Contract DE-AC05-00OR22725. SUPPORTING INFORMATION AVAILABLE Photograph of the TLC-MS interface coupled to the ACQUITY UPLC and 4000 QTRAP systems. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review July 30, 2009. Accepted September 21, 2009. AC901712B (41) Salomonsson, M. L.; Bondesson, U.; Hedeland, M. J. Mass. Spectrom. 2009, 44, 742–754.