Quantification of Farnesylmethylcysteine in Lysates of Peripheral

United Kingdom, Antoni van Leeuwenhoek Hospital/The Netherlands Cancer ... 1066 CX Amsterdam, The Netherlands, and Division of Drug Toxicology, ...
3 downloads 0 Views 98KB Size
Anal. Chem. 2006, 78, 2617-2622

Quantification of Farnesylmethylcysteine in Lysates of Peripheral Blood Mononuclear Cells Using Liquid Chromatography Coupled with Electrospray Tandem Mass Spectrometry: Pharmacodynamic Assay for Farnesyl Transferase Inhibitors Natalie M. G. M. Appels,† Hilde Rosing,† Trevor C. Stephens,‡ Jan H. M. Schellens,§,| and Jos H. Beijnen*,†,§,|

Department of Pharmacy & Pharmacology, Slotervaart Hospital/The Netherlands Cancer Institute, Louwesweg 6, 1066 EC Amsterdam, The Netherlands, AstraZeneca Pharmaceuticals, Alderley Park, Macclesfield, Cheshire SK10 4TG, United Kingdom, Antoni van Leeuwenhoek Hospital/The Netherlands Cancer Institute, Plesmanlaan 121, 1066 CX Amsterdam, The Netherlands, and Division of Drug Toxicology, Department of Biomedical Analysis, Faculty of Pharmaceutical Sciences, Utrecht University, P.O. Box 80082, 3508 TB Utrecht, The Netherlands

Biological effectiveness is an important parameter in determining optimal dosages of molecular targeted drugs, such as farnesyl transferase inhibitors. To determine concentration-effect relationships, robust and quantitative biological assays are a prerequisite. Here, we present a novel assay for protein farnesylation that is based on generation of the biomarker farnesylmethylcysteine (FmC). Quantification was performed with liquid chromatography coupled to tandem mass spectrometry. The assay has been validated based on the most recent FDA guidelines for bioanalytical validation, and all results were within requirements. FmC is formed under the action of an endogenous protease that is activated upon cell lysis. The biomarker could be detected in A549 human lung cancer cells as well as in human peripheral blood mononuclear cells. Incubation of A549 cells with AZD3409, a novel prenyl transferase inhibitor, resulted in a significant decrease of the FmC concentration in the lysates. These findings provide a very good starting point for use of this assay in preclinical and clinical dose finding studies with FTIs. The development of molecular targeted drugs is a rapidly evolving field, especially in cancer treatment. Prenyl transferase inhibitors are a class of molecular targeted drugs that are capable of modulating a number of target proteins including the oncoprotein Ras, via inhibition of the enzymes farnesyl transferase (FTase), geranylgeranyl transferase-1, or both. Ras is an important factor in the signal transduction pathway controlling cell growth. * Corresponding author. Tel: +31 (0) 20 512 4180. Fax: +31 (0) 20 512 4753. E-mail: [email protected]. † Slotervaart Hospital/The Netherlands Cancer Institute. ‡ AstraZeneca Pharmaceuticals. § Antoni van Leeuwenhoek Hospital/The Netherlands Cancer Institute. | Utrecht University. 10.1021/ac051786s CCC: $33.50 Published on Web 03/10/2006

© 2006 American Chemical Society

It exists in three isoforms: H-Ras, N-Ras, and K-Ras that is present in the splice variants Ki4A-Ras and Ki4B-Ras.1 Mutation of Ras at position 12, 13, or 61, which is found in 30% of all human cancers, results in increased activation of cell growth.2 Inhibition of Ras function is possible by inhibition of farnesylation, its first posttranslational modification.3 Several FTase inhibitors (FTIs) have been developed and are currently tested in clinical phase I, II, and III trials.4,5 To determine dose-effect relationships at the biological level, quantitative, well-defined, and reproducible methods are a prerequisite. Most of the available assays determined the farnesylation level of Ras or a surrogate farnesylated protein using semiquantitative western blotting methods. These assays generally are not very robust and the resolution of gel electrophoresis is too low to discriminate between farnesylated and geranylgeranylated proteins (e.g., Ki4B-Ras or RhoB). Recently, novel methods for the quantification of proteins based on the selection of a signature peptide have been reported.6-9 The signature peptide is derived from the protein of interest by specific trypsin cleavage and can be measured with excellent selectivity and high sensitivity using tandem mass spectrometry coupled to liquid chromatography. (1) Barbacid, M. Annu. Rev. Biochem. 1987, 56, 779-827. (2) Bos, J. L. Cancer Res. 1989, 49, 4682-4689. (3) Kato, K.; Cox, A. D.; Hisaka, M. M.; Graham, S. M.; Buss, J. E.; Der, C. J. Proc. Natl. Acad. Sci. U. S. A. 1992, 89, 6403-6407. (4) Crul, M.; de Klerk, G. J.; Beijnen, J. H.; Schellens, J. H. Anticancer Drugs 2001, 12, 163-184. (5) Brunner, T. B.; Hahn, S. M.; Gupta, A. K.; Muschel, R. J.; McKenna, W. G.; Bernhard, E. J. Cancer Res. 2003, 63, 5656-5668. (6) Barnidge, D. R.; Dratz, E. A.; Martin, T.; Bonilla, L. E.; Moran, L. B.; Lindall, A. Anal. Chem. 2003, 75, 445-51. (7) Gerber, S. A.; Rush, J.; Stemman, O.; Kirschner, M. W.; Gygi, S. P. Proc. Natl. Acad. Sci. U. S. A. 2003, 100, 6940-6945. (8) Kuhn, E.; Wu, J.; Karl, J.; Liao, H.; Zolg, W.; Guild, B. Proteomics 2004, 4, 1175-1186. (9) Zhang, F.; Bartels, M. J.; Brodeur, J. C.; Woodburn, K. B. Environ. Toxicol. Chem. 2004, 23, 1408-1415.

Analytical Chemistry, Vol. 78, No. 8, April 15, 2006 2617

Figure 1. Posttranslational modifications of proteins that contain a C-terminal CAAX sequence. First, the protein is farnesylated at the cysteine (C). The second and third posttranslational steps involve proteolysis and carboxymethylation, respectively. The dotted line represents the cleavage point of trypsin. FTase, farnesyltransferase; Rce1, Ras-converting enzyme 1; Icmt, isoprenylcysteine carboxyl methyltransferase.

Figure 2. Chemical structure of FmC.

Ras, as well as other farnesylated proteins, are posttranslationally modified in a cascade of three reactions, i.e., farnesylation at the C-terminal cysteine, proteolysis of the last three C-terminal amino acids, and carboxymethylation of the farnesylated cysteine (Figure 1). It is expected that the intermediates exist only transiently and are rapidly converted to the fully posttranslational modified protein. Thus, a common feature in farnesylated, proteolyzed, and carboxy-methylated proteins is the C-terminal farnesylmethylcysteine (FmC, Figure 2). Examination of a list of known farnesylated proteins10 reveals that incubation with trypsin theoretically results in FmC from H-Ras, Ki4A-, and Ki4B-Ras proteins as well as in other farnesylated proteins such as PRL-3, RhoI, and AGS1 (activator of G-protein signaling). In this paper, we present an absolute quantification method for FmC in lysates of peripheral blood mononuclear cells (PBMCs) as a novel pharmacodynamic assay to determine the biological effects of FTIs. Upon lysis of the PBMCs, FmC was immediately formed under the action of an intracellular protease. FmC was quantitated with LC-MS/MS using a 15N,13C-labeled reference peptide as internal standard. Preclinical proof of principle of the method was obtained in A549 human lung cancer cells treated with the novel dual prenyl transferase inhibitor AZD3409.11 EXPERIMENTAL SECTION Chemicals. [15N,13C3]-Cysteine was purchased at Cambridge Isotope Laboratories (Andover, MA). Methylcysteine, farnesyl bromide, diisopropylethylamine (DIPEA), HEPES, magnesium chloride, and potassium chloride were purchased from Sigma Aldrich (St. Louis, MO). Dimethyl sulfoxide (DMSO) and ammonia (25%) were obtained from Merck (Darmstadt, Germany). Complete protease inhibitor tablets were purchased at Roche Diagnostics (Mannheim, Germany). Peptide sequencing grade dimethylformamide (DMF), HPLC grade acetonitrile was purchased from Biosolve Ltd. (Amsterdam, The Netherlands). Double-distilled water was used throughout the analyses. AZD3409 compound was provided by AstraZeneca Pharmaceuticals (Alderley Park, Cheshire, United Kingdom). (10) Fiordalisi, J. J.; Johnson, R. L.; Weinbaum, C. A.; Sakabe, K.; Chen, Z.; Casey, P. J.; Cox, A. D. J..Biol. Chem. 2003, 278, 41718-41727. (11) Stephens, T. C., Wardleworth, M. J., Matusiak, Z. S., Ashton, S. E., Hancox, U. J.; Bate, M.; Ferguson, R.; Boyle, T. AZD3409, a novel, oral, protein prenylation inhibitor with promising preclinical antitumor activity. Proc. Am. Assoc. Canc. Res. 2003; R4870.

2618

Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

Farnesylation of methylcysteine was performed by dissolving 5.5 mg of the methylcysteine in 0.1 mL of DMF. A 33-µL aliquot of this stock solution was diluted with 635 µL of DMF and 225 µL of DMSO and incubated with 73 µL of DIPEA (0.6 M) and 73 µL of farnesyl bromide (0.2 M).12 After incubation for 1 h at room temperature, the reaction was complete. Stock solutions were stored at nominally 2-8°C. Carboxymethylation and farnesylation of 15N13C3-cysteine was performed by dissolving 3.0 mg of the stable isotopically labeled amino acid in 0.5 mL of water. A 100-µL aliquot of this stock solution was diluted with 900 µL of methanol and lyophilized. A methanolic HCl solution was prepared by dropwise adding 1.6 mL of acetyl chloride to 10 mL of ice-cold, dry methanol.13 The esterification was allowed to proceed for 2 h at room temperature. Solvent was removed by SpeedVac centrifugation. Subsequently, the dry extract was dissolved in 10 µL of water, 72 µL of DMSO, and 185 µL of DMF and incubated with 23 µL of DIPEA (0.6 M) and 23 µL of farnesyl bromide (0.2 M) for 1 h. The stock solution was stored at nominally 2-8 °C. The identity of the synthesized product was confirmed with MS. Instrumentation. Analysis of FmC and its stable isotopically labeled internal standard was carried out using an 1100 Series binary pump with mobile-phase degasser and a cooled autosampler (Agilent Technologies, Palo Alto, CA) coupled on-line to an API2000 triple-quadrupole mass spectrometer equipped with a turbo ion spray source (Sciex, Thornhill, ON, Canada). The analytes were separated on a X-Terra C8 column (100 × 2.1 mm i.d., particle size 5 µm) (Waters Corp., Milford, MA) using a linear gradient from 30 to 95% B in 4 min at a flow of 0.2 mL/min. Eluent A consisted of 10 mM aqueous ammonia in water, and eluent B consisted of acetonitrile. The injection volume was 20 µL. The quadrupoles were operated in the positive ion mode with unit resolution, and the obtained multiple reaction monitoring chromatograms were used for quantification using Analyst software version 1.2 (Sciex). The sensitivity of the mass transitions of m/z 340.3 f 81.3 and 344.3 f 81.3 was optimized for FmC and [15N13C3]-FmC, respectively. Nebulizer and turbogas (both compressed air) were set at 40 and 70 psi, respectively, while the curtain gas and collision gas (both N2) were operated at 40 psi and 1.04 × 1015 molecules/cm2, respectively. The ion spray voltage was kept at 5500 V, with a source temperature of 390 °C. The dwell time was 250 ms. Clinical Sample Preparation. Whole blood samples of 20 mL (containing ∼3.3 × 106 cells) from six individuals were collected in heparinized tubes for the determination of the basal FmC concentration in lysates of PBMCs. Within 30 min of collection, the blood was transferred to an Accuspin tube (Sigma) and centrifuged for 10 min at 1000g and room temperature. The PBMC-containing fraction in the upper chamber of the Accuspin tube was transferred to a fresh polypropylene tube. Phosphatebuffered saline (PBS) was added to a final volume of 50 mL. After centrifugation for 15 min at 300g, the cell pellet was washed with 5 mL of sterile distilled water for 5 s to lyse contaminating red blood cells. Subsequently, PBS was added to a final volume of 40 (12) Pompliano, D. L.; Gomez, R. P.; Anthony, N. J. J. Am. Chem. Soc. 1992, 114, 7945-7946. (13) Ficarro, S. B.; McCleland, M. L.; Stukenberg, P. T.; Burke, D. J.; Ross, M. M.; Shabanowitz, J.; Hunt, D. F.; White, F. M. Nat. Biotechnol. 2002, 20, 301-305.

mL and the sample was centrifuged for 15 min at 300g. This was repeated 2 times. After the last centrifugation step, the cell pellet was stored at -70 °C until analysis. Sample Processing. Frozen pellets of PBMCs were lysed with 100 µL of ice-cold hypotonic buffer (10 mM HEPES, pH 7.2, 1.5 mM MgCl2, 10 mM KCl) and vigorously mixed. After centrifugation for 10 min at 23100g and 4 °C, 50 µL of clear lysate was added to an ice-cold mixture of 50 µL of acetonitrile and 10 µL of the working solution of the internal standard. The samples were briefly vortexed for 10 s and then centrifuged for 10 min at 23100g and 4 °C. The supernatant was transferred to a glass autosampler vial, and an aliquot of 20 µL was injected onto the analytical column. The calibration standards were processed by adding 50 µL of hypotonic buffer instead of acetonitrile. Preparation of Calibration Standards and Quality Control Samples. Two sets of FmC stock solutions were independently prepared at a concentration of 10 mM. One solution was diluted with acetonitrile to obtain working solutions with concentrations of 100 and 1000 nM, respectively. The FmC working solutions were spiked into acetonitrile to obtain calibration standards ranging from 0.249 to 14.9 nM. The second stock solution was also diluted with acetonitrile to obtain working solutions with concentrations of 1500, 300, and 150 nM, respectively. These working solutions were spiked into plasma that was diluted 10 times with hypotonic buffer to provide quality control (QC) samples at three concentration levels named QC LOW, QC MID, and QC HIGH (with concentrations of 0.794, 2.38, and 11.9 nM, respectively). The stock solution of the internal standard was diluted with acetonitrile to yield a working solution with a theoretical concentration of ∼200 nM [15N13C3]-FmC. All solutions and calibration samples were stored at nominally 2-8 °C. The quality control samples were prepared freshly in an ice bath for each analytical run. Validation Procedure. A full validation program based on the FDA guidelines was executed for the assay of FmC in lysates of human PBMCs.14,15 Linearity. Eight nonzero plasma calibration standards, with FmC concentrations ranging from 0.249 to 14.9 nM, were prepared and analyzed in duplicate in three separate analytical runs. The calibration curves were calculated by least-squares linear regression using a weighting factor of 1/x2 (the reciprocal of the squared concentration). The standard concentrations were back-calculated from a constructed calibration curve. The deviation from nominal concentrations over three runs should be within (15% ((20% at the lower limit of quantitation) for the calibration standards. Accuracy and Precision. The accuracy and precision of the assay were established by analyzing QC samples of FmC. Five replicates of each sample were analyzed together with a complete set of calibration standards in three analytical runs. The intraassay accuracy was determined as the percent difference between the mean concentration per analytical run and the nominal concentration. The interassay accuracy was determined as the percent difference between the mean concentration after three analytical (14) U.S. Food and Drug Administration, Center for Drug Evaluation and Research, Guidance for Industry: Bioanalytical Method Validation. 2001; www.fda.gov/cder/guidance/4252fnl.html. (15) Rosing, H.; Man, W. Y.; Doyle, E.; Bult, A.; Beijnen, J. H. J. Liq. Chromatogr. Relat. Technol. 2000, 23, 329-254.

runs and the nominal concentration.The coefficient of variation provided the measure of intra- and interassay precision. The accuracy and precision should both be within 15%. Ion Suppression and Recovery. The ion suppression of FmC by matrix constituents was determined by comparing the areas of blank diluted plasma samples that were processed, evaporated to dryness, and reconstituted with working solutions of FmC to the areas of unprocessed plasma-free samples at the same concentration levels. Both sets of samples were prepared with the same working solutions. The recovery after protein precipitation was determined by comparing the chromatographic peak areas of processed plasma quality control samples to a corresponding set of processed blank samples, which were evaporated to dryness and reconstituted with working solutions of FmC. Specificity and Selectivity. Sets of drug-free plasma (double blank, blank, and QC LOW samples) from four individual donors were diluted with hypotonic buffer, prepared for analysis, and analyzed to determine whether endogenous plasma constituents interfere with the analyte or internal standard. Interference may occur when coeluting endogenous compounds produce ions at the same m/z values that are used to monitor the analyte and internal standard. The peak areas of endogenous compounds coeluting with the analyte should not exceed 20% of the analyte peak area at the QC LOW level. Stability. The stability of FmC was investigated at various concentrations during all steps of the analysis, which includes the stability in the stock solutions (10 mM), in acetonitrile (final matrix of the calibration standards), in 10 times diluted plasma (final matrix of the QC samples) after one freeze (-20 °C)-thaw cycle, in PBMC lysate during sample preparation, and in the final extract. One to three levels per condition were tested. FmC is considered stable in the biological matrix when 85-115% of the initial concentration is found. In stock solutions, FmC is considered stable when 95-105% of the initial concentration is found. Pharmacodynamic Studies. A549 human lung cancer cells were seeded in 14-cm Petri dishes and treated with AZD3409 on day 2. After 3 days, cells were scraped and the pellets were stored at -70 °C. The cell pellets were lysed with hypotonic buffer and centrifuged for 10 min at 23000g and 4 °C. The lysates were immediately analyzed. The protein content was determined with a Bradford assay. An IC50 curve was constructed with Graphpad Prism software (version 4.02 for Windows, GraphPad Software, San Diego, CA). The minimal FmC concentration was set at 0 pmol/mg of protein. The steepness of the slope described by the Hill factor, the IC50, and the maximal FmC concentration were estimated by the software. RESULTS AND DISCUSSION Method Development. We synthesized FmC by chemical farnesylation of methylcysteine and used it as a reference standard. Figure 3 shows the positive ion mass spectrum of ∼150 µM FmC (mass range from 100 to 400 amu). The protonated molecule was observed at m/z 340.3. The spectrum also showed a large peak at m/z 130.5, which could be assigned to the DIPEA that was used as a catalyst for the farnesylation reaction. The other peaks probably represent impurities in the reaction solvents. The MH+ peak of the internal standard, [15N13C3]-FmC, was observed at m/z 344.3 (data not shown). Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

2619

Figure 3. Positive ion Q1 spectrum of FmC (MH+, m/z 340.3), recorded with an API2000 from m/z 100 to 400. A solution of ∼150 µM in a solution of methanol/1% formic acid (80:20, v/v) was continuously infused at 10 µL/min. The peak at m/z 130.5 results from DIPEA that was used in the synthesis of FmC.

Figure 5. (A) Representative SRM chromatogram of 1.0 nM FmC in hypotonic buffer/acetonitrile (50:50, v/v). (B) SRM chromatogram of a clinical sample containing 0.312 nM FmC in the final extract. A whole blood sample of 20 mL was processed to obtain PBMCs, as described in the text.

Figure 4. (A) Product ion scan of FmC recorded in Q3 of an API2000 mass spectrometer after selection in Q1 of the precursor ion at m/z 340.3 and induction fragmentation with nitrogen gas. The product ion scan of 15N13C3-FmC showed a similar pattern, except for the peak at m/z 135.9, which shifted to m/z 140.0. This peak was assigned to methylcysteine. The most intense peak at m/z 81.3 was chosen for multiple reaction monitoring. (B) The potential pathway for generation of this fragment is presented by three cleavages I, II, and III.

In the product ion mass spectrum of both FmC and [15N13C3]FmC, the most intense fragment peak was observed at m/z 81.3 as shown in Figure 4A. Generation of the same fragment of both FmC and [15N13C3]-FmC indicates that the fragment is formed in the farnesyl moiety. The proposed fragmentation scheme is depicted in Figure 4B. First, cleavage of the farnesyl moiety occurs resulting in two fragments, which can be observed at m/z 135.9 ([M + H+] of methylcysteine) or at m/z 204.9 ([M + H+] of farnesyl) depending on which fragment acquired the positive charge. Subsequently, 4-methyl-1,4-hexadiene ([M + H+] at m/z 95.0) and 2-methyl-2,6-octadiene ([M + H+] at m/z 108.9) are formed after cleavage of the second double bond. Finally, the peak at m/z 81.3 corresponds to elimination of the C6-methyl group from the 4methyl-1,4-hexadiene fragment. For quantitative detection of FmC and the internal standard, the most abundant mass transitions from the protonated molecules (m/z 340.3 and 344.3, respectively) to the 4-methyl-1,4-pentadiene fragment of m/z 81.3 were chosen. FmC is a very hydrophobic compound, difficult to elute from a C18 reversed-phase column. Therefore, a X-Terra C8 column was used for the chromatography of FmC. A steep gradient from 2620 Analytical Chemistry, Vol. 78, No. 8, April 15, 2006

30 to 95% acetonitrile in 4 min was used to diminish peak broadening. Several mobile-phase additives were tested in flow injection analysis experiments, and the addition of aqueous ammonia to the eluent resulted in the highest signal-to-noise ratios for the analyte in the positive ionization mode. To avoid nonspecific binding of FmC to the tubes, all samples were prepared in 50% acetonitrile or in plasma that had been diluted 10 times to a protein concentration of ∼5 mg/mL. Figure 5 shows a SRM chromatogram of FmC. The retention time was ∼9 min. Our next step in the development was optimization of the digestion method. A reference H-Ras protein that did not contain the three C-terminal amino acids and was both farnesylated and methylated at the C-terminal cysteine was not available. Therefore, we lysed pellets of A549 human lung cancer cells and analyzed FmC before and after incubation with an excess of trypsin. Surprisingly, FmC could be detected in the lysates before digestion and its peak decreased after incubation with trypsin. To investigate whether FmC was present within the cell cytosol or formed upon disruption of the cells, we tested different lysis buffers with increasing concentrations of protease inhibitors. The peak area of FmC decreased significantly in the presence of protease inhibitors. Apparently, the formation of FmC could be arrested through the inhibition of intracellular proteases during the lysis process. This implies that cleavage of FmC from the farnesylated protein is due to a protease that was not active in the intact cell. Identification of this endogenous protease has not been performed yet. We assumed that cleavage by this enzyme does not depend on the amino acid at the N-terminal site of the farnesylated cysteines, which means that FmC is derived not only from farnesylated H-Ras proteins but theoretically from any protein present in the cytosol that has been farnesylated and fully posttranslational modified.

Table 1. Intra- and Interassay Performance Data for the Analysis of FmC in Human Plasma That Has Been Diluted Ten Times with Hypotonic Buffera run 1 2 3

nominal concn (nM)

mean calcd concn (nM)

accuracy (% bias)

precision (% CV)

0.794 0.794 0.794

0.840 0.722 0.752

5.77 -9.12 -5.26

5.34 6.37 8.82

-2.87

9.28

-0.336 -2.69 -0.504

6.71 8.12 2.43

-1.18

5.86

-7.56 -10.1 -4.54

4.41 3.30 5.45

-7.39

4.89

interassay Figure 6. FmC concentration in lysates of A549 cells and human PBMCs. The FmC concentrations were corrected for the total protein concentrations. Lysates 1-3 represent A549 cells; lysates 4-9 represent PBMCs of six human volunteers. The coefficient of variation was 64%.

Therefore, FmC might be used as a biomarker for the total level of farnesylated proteins. FmC could be detected in lysates of both cultured tumor cells (A549) and human PBMCs as shown in Figure 6. The FmC concentration was corrected for the total protein concentration and expressed in picomoles per milligram of protein. The basal levels in PBMC lysates varied from 0.18 to 1.38 pmol of FmC/ mg of protein. The variation coefficient was 64% (n ) 6). Nevertheless, the presence of detectable concentrations of FmC in both PBMCs and tumor cells provides an excellent starting point for use of this assay as a surrogate pharmacodynamic end point in preclinical and clinical studies with FTIs. Validation. Linearity. Duplicate calibration standards, prepared in acetonitrile with a dynamic range of 0.249-14.9 nM, were analyzed in three analytical runs. The best fit for the calibration curves was obtained by using a weighting factor of the reciprocal of the squared concentration. The correlation coefficients were 0.995 or better. The calibration standards were back-calculated from the responses. The deviations from the nominal concentration were between -3.61 and 7.11% for all concentrations with precisions less than 9.02%. Accuracy and Precision. PBMC lysate contains endogenous FmC and could not be used as matrix to prepare validation samples. Therefore, plasma was diluted 10 times with hypotonic buffer to obtain protein levels that were equal to the levels in PBMC lysates, i.e., ∼5 mg/mL. The diluted plasma solution was used to prepare the QC samples. The intra- and interassay performance data are presented in Table 1. The measured intraassay accuracies ranged from -10.1 to 5.8%, while the interassay accuracies were within (7.4%. The mean intraassay precision did not exceed 8.8%. The interassay precision was less than 7.3% for all tested concentrations. These results prove that the FmC assay fulfills the performance criteria for bioanalytical methods. Furthermore, the nonspecific binding of FmC to the tubes is prevented by proteins at a concentration of ∼5 mg/mL. Ion Suppression and Recovery. The mean ion suppression of FmC was -4.1(10.4%, indicating that there was no significant ion suppression. Also, for the internal standard, hardly any ion suppression was detected. However, it was observed that the ion suppression increased after multiple injections of plasma or lysate samples. Therefore, a blank sample (10 mM aqueous ammonia/acetonitrile, 50:50, v/v) was injected after each sample, except for the calibration standards. Protein precipitation recovery of FmC was 91.7 ( 3.6%. The internal standard also had a recovery of 91.7%.

1 2 3

2.38 2.38 2.38

2.37 2.32 2.37

interassay 1 2 3

11.9 11.9 11.9

11.0 10.7 11.4

interassay a

Number of replicates, five throughout.

Specificity and Selectivity. The selectivity of the method was established by the analysis of double blank, blank, and QC LOW samples prepared in four individual batches of human plasma that were diluted 10 times with hypotonic buffer. No interferences from endogenous material at the retention time of FmC with areas