Targeted Quantitation of Overexpressed and Endogenous Cystic

Nov 30, 2009 - Cystic fibrosis transmembrane conductance regulator (CFTR) functions as an ion channel in the apical plasma membrane of epithelial cell...
0 downloads 0 Views 545KB Size
Download PDF
Anal. Chem. 2010, 82, 336–342

Targeted Quantitation of Overexpressed and Endogenous Cystic Fibrosis Transmembrane Conductance Regulator Using Multiple Reaction Monitoring Tandem Mass Spectrometry and Oxygen Stable Isotope Dilution Hui Jiang, Alexis A. Ramos, and Xudong Yao* Department of Chemistry, University of Connecticut, Storrs, Connecticut 06269 Cystic fibrosis transmembrane conductance regulator (CFTR) functions as an ion channel in the apical plasma membrane of epithelial cells. Mutations in the gene coding for CFTR cause cystic fibrosis (CF). A major cellular dysfunction is insufficient apical plasma membrane expression of the protein. Its correction is important for developing new CF therapeutics and treatments, which requires a sensitive and precise method for quantifying apical plasma membrane CFTR. We report the first method of liquid chromatography-tandem mass spectrometry for quantifying endogenous and overexpressed CFTR in HT29 and BHK cells. For low level of endogenous CFTR from HT29, the target protein in the cell lysate was enriched by immunoprecipitation using antiCFTR antibody MAB3484 or M3A7. For overexpressed CFTR from BHK, the cell lysate prepared by differential detergent fractionation or surface biotinylation was used directly without immunoprecipitation. Proteins in the enriched CFTR preparations or cell lysates were digested with proteases, and a surrogate marker peptide designated as CFTR01 (NSILTETLHR) was successfully quantified using the method of multiple reaction monitoring and stable isotope dilution with an 18O-labeled reference peptide (CFTR01-18O4) as the internal standard. CFTR quantified in this work ranged from a few tens of picograms to low nanograms per million of cells. Cystic fibrosis transmembrane conductance regulator (CFTR) is a member of ATP-binding cassette (ABC) transporter superfamily and functions as Cl- channel in the apical membrane of epithelial cells. Mutations in the CFTR gene cause cystic fibrosis (CF) with clinical symptoms including chronic lung disease,pancreaticdysfunction,andelevatedsweatelectrolytes.1-3 The common CFTR mutant with a single phenylalanine residue deletion (F508del-CFTR) accounts for the major CF population (http://www.genet.sickkids.on.ca/cftr). This mutant is prone to misfolding and thus being retained inside the cell by the endo* To whom correspondence should be addressed. E-mail: [email protected]. Phone: 1-860-486-6644. (1) Riordan, J. R. Annu. Rev. Biochem. 2008, 77, 701–726. (2) Guggino, W. B.; Stanton, B. A. Nat. Rev. Mol. Cell Biol. 2006, 7, 426–436. (3) Amaral, M. D. J. Mol. Neurosci. 2004, 23, 41–48.

336

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

plasmic reticulum and degraded by the proteasome,3,4 which results in very little or no localization of the protein in the apical plasma membrane to maintain normal Cl- channel activity. Thus, new therapeutics for rescuing F508del-CFTR to the cell surface is being actively pursued.5,6 In order to evaluate the efficacy of lead compounds and drug candidates, an efficient method is much needed to detect and quantify the apical plasma membrane CFTR expression. The common method for quantifying CFTR is Western analysis,7 but it does not have sufficient reproducibility for robust and highthroughput applications for the drug development. CFTR aggregation and degradation and the missing “perfect” CFTR antibody8 are among the analytical challenges. Tandem mass spectrometry (MS/MS), representatively multiple reaction monitoring (MRM) or selected reaction monitoring (SRM) method, has recently attracted a lot of attention for quantitation of targeted proteins in complex biomatrixes.9-17 The sensitivity and specificity afforded by the MS/MS measurement lift the dependence on high-quality antibodies for conventional (4) Farinha, C. M.; Amaral, M. D. Mol. Cell. Biol. 2005, 25, 5242–5252. (5) Powell, K.; Zeitlin, P. L. Adv. Drug Delivery Rev. 2002, 54, 1395–1408. (6) Amaral, M. D.; Kunzelmann, K. Trends Pharmacol. Sci. 2007, 28, 334– 341. (7) Farinha, C. M.; Penque, D.; Roxo-Rosa, M.; Lukacs, G.; Dormer, R.; McPherson, M.; Pereira, M.; Bot, A. G.; Jorna, H.; Willemsen, R.; Dejonge, H.; Heda, G. D.; Marino, C. R.; Fanen, P.; Hinzpeter, A.; Lipecka, J.; Fritsch, J.; Gentzsch, M.; Edelman, A.; Amaral, M. D. J. Cystic Fibrosis 2004, 3 (2), 73–77. (8) Farinha, C. M.; Mendes, F.; Roxo-Rosa, M.; Penque, D.; Amaral, M. D. Mol. Cell. Probes 2004, 18, 235–242. (9) Carr, S. A.; Anderson, L. Clin. Chem. 2008, 54, 1749–1752. (10) Sherman, J.; McKay, M. J.; Ashman, K.; Molloy, M. P. Proteomics 2009, 9, 1120–1123. (11) Duncan, M. W.; Yergey, A. L.; Patterson, S. D. Proteomics 2009, 9, 1124– 1127. (12) Lange, V.; Malmstrom, J. A.; Didion, J.; King, N. L.; Johansson, B. P.; Schafer, J.; Rameseder, J.; Wong, C. H.; Deutsch, E. W.; Brusniak, M. Y.; Buhlmann, P.; Bjorck, L.; Domon, B.; Aebersold, R. Mol. Cell. Proteomics 2008, 7, 1489–1500. (13) Wolf-Yadlin, A.; Hautaniemi, S.; Lauffenburger, D. A.; White, F. M. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 5860–5865. (14) Keshishian, H.; Addona, T.; Burgess, M.; Kuhn, E.; Carr, S. A. Mol. Cell. Proteomics 2007, 6, 2212–2229. (15) Gerber, S. A.; Rush, J.; Stemman, O.; Kirschner, M. W.; Gygi, S. P. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 6940–6945. (16) Barnidge, D. R.; Goodmanson, M. K.; Klee, G. G.; Muddiman, D. C. J. Proteome Res. 2004, 3, 644–652. (17) Anderson, L.; Hunter, C. L. Mol. Cell. Proteomics 2006, 5, 573–588. 10.1021/ac902028f  2010 American Chemical Society Published on Web 11/30/2009

biochemical quantitation of protein targets. MRM monitors gasphase dissociation reactions of target analytes, which requires the sequential detection of an analyte precursor ion as the reactant followed by one or several analyte fragment ions as the products; therefore, it is highly specific and can be applied to complex samples with minimal component separation. The specificity can be further increased when MRM-MS/MS is combined with liquid chromatography (LC) and stable isotope dilution.14-17 Furthermore, MRM approaches for protein quantitation mitigate sample preparation difficulties of membrane proteins. Protein samples are digested into peptides that are relatively easy to prepare for mass spectrometry (MS). Selected tryptic peptides of target proteins are commonly used as surrogate markers for MS-based protein quantitation. This mitigation is a very significant analytical advantage, especially for membrane proteins like CFTR that are easy to aggregate and degrade. Furthermore, multiplexing potential of MRM methods allows simultaneous quantification of many analytes in one particular sample, which has high potential in both basic and clinical proteomic analysis.12-15,17 Herein, we report the first MS assay for CFTR quantification. A tryptic CFTR peptide is selected as the marker peptide for the protein, and its amount in digests of CFTR-enriched protein samples, prepared by differential detergent fractionation, surface biotinylation, or immunoprecipitation, is quantified using LC-MRM-MS/MS and an internal standard of oxygen isotope-labeled reference marker peptide. EXPERIMENTAL SECTION Materials. Digitonin, Triton X-100, Trizma base, 1,4-piperazinediethanesulfonic acid (PIPES), sucrose, phenylmethylsulfonyl fluoride (PMSF), methotrexate, dimethyl pimelimidate dihydrochloride (DMP), urea, iodoacetamide (IAA), trifluoroacetic acid (TFA), protease inhibitor cocktail, and Protein G immunoprecipitation kit were bought from Sigma-Aldrich (St. Louis, MO). NuPAGE 4-12% bis-tris gel, NuPAGE MOPS SDS running buffer, Invitrolon PVDF filter paper sandwich (0.45 µm), WesternBreeze chromogenic immunodetection system, and Quant-iT protein assay kit (for Qubit fluorometer) were purchased from Invitrogen (Carlsbad, CA). Anti-CFTR monoclonal antibodies M3A7 and MAB3484 (L12B4) were purchased from Millipore through Fisher Scientific (Pittsburgh, PA). Dithioerythritol (DTE) and β-mercaptoethanol (β-ME) were bought from Fisher Scientific (Pittsburgh, PA). Lys-C and trypsin were obtained from Roche Applied Sciences (Indianapolis, IN). SilverSNAP stain and cell surface protein isolation kit were purchased from Pierce, and 18Owater (>97%) was purchased from Cambridge Isotope Laboratories (Andover, MA) or obtained as a gift from Olinax (Hamilton, ON, Canada). Alkaline phosphatase, Calf intestinal phosphatase (CIP), was bought from New England Biolabs (Ipswich, MA). Marker peptide CFTR01 was custom-synthesized by AnaSpec (San Jose, CA). Baby hamster kidney (BHK) cells and human intestinal HT29 cells were cultured by Dr. T. Smith of the Cell Culture Facility at the University of Connecticut, using standard monolayer plate culture protocols. Purified full-length CFTR samples were gifts from Dr. L. J. DeLucas at University of Alabama at Birmingham, Dr. J. R. Riordan at University of North Carolina, and Dr. J. He at Accelagen (San Diego, CA).

Cell Lysis and Protein Extraction. For differential detergent fractionation experiment, CFTR overexpressed in BHK cells was processed sequentially with digitonin extraction buffer (10 mM PIPES, 0.015% digitonin, 300 mM sucrose, 100 mM NaCl, 3 mM MgCl2, 5 mM EDTA, 1 mM PMSF) and Triton X-100 lysis buffer (25 mM Tris pH 7.4, 1.0% Triton X-100, 150 mM NaCl, 1% protease inhibitor cocktail) using a similar procedure to published protocols.18,19 HT29 cells were treated with Triton X-100 lysis buffer only, without prior digitonin extraction. The obtained protein fractions were kept at -20 °C for further analysis after their total protein content was measured with the Quant-iT assay kit. Immunoprecipitation. Immunoprecipitation was performed according to published procedures20,21 and manufacturer instructions using a Protein G immunoprecipitation kit. Purified fulllength CFTR was used to develop an immunoprecipitation protocol first. A buffer solution of 5 µg of monoclonal CFTR antibody MAB3484 or M3A7 was incubated with 50 µL of Protein G beads. After centrifugation, the flow-through was collected for later analysis. Then 2.1 µg of purified CFTR was added for incubation with the antibody-bound Protein G beads overnight at 4 °C in a thermomixer. Again, after centrifugation, the flow-through was collected. All of the proteins captured on the beads were then eluted out using glycine buffer (pH 2.8), and the pH value of the eluent solution was adjusted to neutral with 1 M Tris buffer (pH 8.0). Anti-CFTR antibodies were also cross-linked to Protein G beads using DMP. Cell lysate fractions were then processed similarly and the total protein contents of the eluate solutions were measured with Quant-iT assay kit. Surface Biotinylation. Surface biotinylation22-24 was performed on BHK cells using the surface protein isolation kit from Pierce, following the manufacturer instructions. Briefly, cells were first labeled with a thiol-cleavable amine-reactive biotinylation reagent, sulfosuccinimidyl-2-(biotinamido)ethyl-1,3-dithiopropionate (sulfo-NHS-SS-biotin) and subsequently lysed with a mild detergent. The labeled proteins were captured with avidin agarose and released using Triton X-100 buffer containing 50 mM dithiothreitol (DTT). The total protein content of the sample was quantified with Quant-iT assay kit. SDS-PAGE and Western Blot Analysis. SDS-PAGE was run with NuPAGE 4-12% bis-tris gel using MOPS buffer and detected with SilverSNAP staining kit. For Western blot, the gel was transferred to Invitrolon PVDF membrane and detected with WesternBreeze chromogenic immunodetection kit. CFTR01-18O4 Preparation. CFTR01 (95 nmol) was dissolved in 50 µL of H218O (>97%) in a tube on ice. A solution of 1 µL of TFA was added to 49 µL of H218O (>97%) in another tube on ice. Then the two solutions were mixed and incubated at 37 °C in a thermomixer. The reaction was monitored by a (18) Ramsby, M. L.; Makowski, G. S. Methods Mol. Biol. 1999, 112, 53–66. (19) McCarthy, F. M.; Burgess, S. C.; van den Berg, B. H.; Koter, M. D.; Pharr, G. T. J. Proteome Res. 2005, 4, 316–324. (20) Seavilleklein, G.; Amer, N.; Evagelidis, A.; Chappe, F.; Irvine, T.; Hanrahan, J. W.; Chappe, V. Am. J. Physiol.: Cell Physiol. 2008, 295, C1366–C1375. (21) Wang, X.; Venable, J.; LaPointe, P.; Hutt, D. M.; Koulov, A. V.; Coppinger, J.; Gurkan, C.; Kellner, W.; Matteson, J.; Plutner, H.; Riordan, J. R.; Kelly, J. W.; Yates, J. R., III.; Balch, W. E. Cell 2006, 127, 803–815. (22) Mohamed, A.; Ferguson, D.; Seibert, F. S.; Cai, H. M.; Kartner, N.; Grinstein, S.; Riordan, J. R.; Lukacs, G. L. Biochem. J. 1997, 322 (Pt 1), 259–265. (23) Jang, J. H.; Hanash, S. Proteomics 2003, 3, 1947–1954. (24) Luo, Y.; McDonald, K.; Hanrahan, J. W. Biochem. J. 2009, 419, 211219 ff.

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

337

Q-TOFmicro mass spectrometer (Waters, Milford, MA) with MassLynx 4.1 software. After 85 h, the mixture was divided into aliquots, frozen, and lyophilized. Stock solution of this 18Olabeled reference marker peptide was prepared in water. Protein Digestion. Purified CFTR, cell lysates, and samples prepared by immunoprecipitation or surface biotinylation were reduced with DTE, alkylated with IAA, and digested with proteases according to standard procedures, except for enzyme-toprotein ratios. After the excess IAA was reacted with βME, sequential digestion with Lys-C at ∼4 M urea concentration and trypsin at ∼1 M urea concentration was performed. If the total protein content was low, e.g., for purified CFTR or CFTR eluate after immunoprecipitation, the enzyme-to-protein ratio used was between 1:5 and 1:10. If the total protein content was high, e.g., for cell lysate, the enzyme-to-protein ratio used was between 1:35 and 1:50. After digestion, peptides of purified CFTR sample were subject to LC-MS measurement directly for marker peptide selection. For all of the other samples, reference marker peptide (final concentration of 1.0 fmol/µL) was added before LC-MS/ MS measurement as internal standard for quantitation. LC-MS/MS Measurements. High-performance liquid chromatography (HPLC) was performed on a Shimadzu 10ADvp system: solvent A, FA/ACN/H2O ) 2:10:988 (v/v/v); solvent B, FA/H2O/ACN ) 2:10:988 (v/v/v); column temperature, 60 °C. Separation of digestion mixture was carried out using a reversed-phase column (Hypersil GOLD, 1.9 µm, 100 mm × 1.0 mm, Thermo Scientific, Waltham, MA). Data-dependent acquisition (DDA) MS/MS of tryptic digests of purified CFTR was performed on Q-TOFmicro (Waters, Milford, MA) with MassLynx 4.1 software. Peptide search was done using Protein Global Server 2.0 software. MRM-MS/MS was carried out on 4000 QTrap (Applied Biosystems, Foster City, CA) with Analyst 1.4.2 software for digests of CFTR-containing samples. Transitions for marker peptide CFTR01 (592.3/869.5) and reference peptide CFTR01-18O4 (596.3/877.5) were monitored simultaneously, 250 ms each. Q1 and Q3 resolutions were set to be unit. Typical MS instrument parameters were the following: CUR, 20.00; IS, 5500.00; TEM, 300.00; GS1, 20.00; GS2, 40.00; ihe, OFF; CAD, 12.00; DP, 90.00; EP, 10.00; CE, 32.50; CXP, 30.00. Various concentrations of CFTR01 (0, 0.25, 0.5, 1.0, 2.5, and 5.0 fmol/µL) and 1.0 fmol/µL of CFTR01-18O4 were mixed in aqueous solution with 10% methanol and 0.1% TFA. Data obtained were used to establish calibration curves. RESULTS AND DISCUSSION CFTR Enrichment. Two types of cells were used for the current study: transfected BHK cells that stably overexpressed CFTR when treated with methotrexate and HT29 cells that expressed endogenous CFTR. Cell lysate samples enriched with CFTR were prepared by three different methods and used to develop the MS assay: differential detergent fractionation, immunoprecipitation, and surface biotinylation. With the use of a series of detergents, differential detergent fractionation sequentially extracted proteins from cells.18,19 The digitonin and Triton X-100 fractions of BHK cells were used directly in this study, because the cells overexpressed CFTR. Digitonin is known to interact with cholesterol in the cell plasma 338

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

Figure 1. Western blot of CFTR eluate after immunoprecipitation on antibody-cross-linked Protein G beads, along with cell lysates and purified CFTR: lane 1, immunoprecipitation eluate corresponding to 133 µL of HT29 Triton X-100 lysate (equivalent to 9.2 × 106 cells) with MAB3484; lane 2, immunoprecipitation eluate corresponding to 133 µL of HT29 Triton X-100 lysate with M3A7; lane 3, 10 µL of HT29 Triton X-100 lysate (equivalent to 0.69 × 106 cells); lane 4, immunoprecipitation eluate corresponding to 14 µL of BHK Triton X-100 lysate (equivalent to 0.62 × 106 cells) with MAB3484; lane 5, immunoprecipitation eluate corresponding to 14 µL of BHK Triton X-100 lysate with M3A7; lane 6, 10 µL of BHK Triton X-100 lysate (equivalent to 0.44 × 106 cells); lane 7, immunoprecipitation eluate corresponding to 32 µL of BHK digitonin lysate (equivalent to 1.5 × 106 cells) with MAB3484; lane 8, immunoprecipitation eluate corresponding to 32 µL of digitonin lysate with M3A7; lane 9, 10 µL of BHK digitonin lysate (equivalent to 0.46 × 106 cells); lane 10, 25 ng of purified CFTR.

membrane to form pores and extract soluble proteins from the cytosol; Triton X-100 solubilizes plasma membrane and organelle proteins.18,19 In order to enrich low level of endogenous CFTR from cell lysates, an immunoprecipitation protocol was developed, using immobilized anti-CFTR monoclonal antibodies MAB3484 and M3A7 on Protein G beads and purified full-length CFTR samples. As shown in Figure S1 in the Supporting Information, both MAB3484 and M3A7 were able to bind Protein G beads to enrich CFTR in immunoprecipitation buffer successfully; both antibodies and CFTR showed bands mainly in the eluate (bound protein fraction, lanes 3 and 6) but little in the flow-through (unbound protein fraction). Not surprisingly, in addition to CFTR, antibodies were also released from the beads under the elution condition when they were not covalently linked to the beads. Antibodies were also cross-linked to the beads with DMP following a published procedure21 before CFTR immunoprecipitation was performed. As expected, after cross-linking, antibody was not observed in the CFTR eluate and CFTR immunoprecipitation still worked sufficiently (Supporting Information Figure S2). Immunoprecipitation of cell lysates was carried out using Protein G beads cross-linked with a CFTR antibody. Figure 1 showed the Western blot for CFTR eluate after immunoprecipitation, along with cell lysates and purified CFTR as control. For 10 µL of HT29 Triton X-100 lysate (equivalent to 0.69 × 106 cells), no CFTR was observed (lane 3). After immunoprecipitation with either MAB3484 (lane 1) or M3A7 (lane 2), CFTR was significantly enriched, enabling the detection of low level of endogenous CFTR in HT29 cells. For BHK cells overexpressing CFTR, the eluate after immunoprecipitation with either MAB3484 (lane 4) or M3A7 (lane 5) showed comparable band to 10 µL of Triton X-100 lysate (equivalent to 0.44 × 106 cells, lane 6), further confirming the effectiveness of the immunoprecipitation enrichment of CFTR. The smear on lanes 4, 5, and 6 was likely due to CFTR aggregation, similar to purified CFTR in lane 10. As for the digitonin fraction, similar to HT29 cells, 10 µL of

Figure 2. (A) LC-MS total ion chromatogram for digest of purified CFTR. (B) Identified peptides shown in red, covering 61% of the CFTR sequence (complete sequence coverage for the analysis was not further pursued, due to the fact that the purpose of this work was to identify high-response marker peptides). Marker peptide CFTR01 (NSILTETLHR) is italicized and underlined.

lysate (equivalent to 0.46 × 106 cells) did not show observable CFTR band (lane 9), whereas immunoprecipitation with either antibody (lane 7 or 8) enriched CFTR efficiently. Apical plasma membrane CFTR from BHK cells was enriched using surface biotinylation. The biotinylation reagent sulfo-NHSSS-biotin had a reactive group for modifying amine groups on proteins, a biotin moiety that allowed affinity enrichment of the reagent-modified proteins, and a disulfide bond that allowed reductive releasing of the reagent-modified proteins from immobilized avidin without eluting endogenous biotinyl proteins. It should be noted that, although surface biotinylation methods are commonly used to enrich plasma membrane CFTR for Western analysis,24 there are only three theoretical extracellular amino groups on the Lys114, Lys329, and Lys892 side chains, which are available for modification (CFTR_HUMAN, UniProtKB/SwissProt: P13569). In addition, it was important to include a mild detergent, e.g., Triton X-100, in the elution buffer for efficient recovery of the biotinylated CFTR; the detergent type and concentration needed to have low detrimental effects on the subsequent LC separation and MS analysis. Selection of Marker Peptide CFTR01. Tryptic digest of purified full-length CFTR was analyzed by LC-MS (Figure 2A). Selection of a candidate peptide as the surrogate protein marker was based on peptide performance during LC-MS/MS analysis

(Figure 2A). The marker peptide selection was facilitated by the availability of the purified full-length CFTR protein, which was used to produce tryptic peptides. Figure 2A showed the total ion chromatogram for CFTR peptides. Protein database search based on DDA MS/MS data indicated a 61% sequence coverage as shown in Figure 2B (identified peptides are in red). Common criteria for selecting marker peptide candidates for MRM-MS/MS analysis were considered, including peptide uniqueness, peptide length, MS ionization efficiency, and biased MS/ MS signal.17,25 Against these standards, examination of peptide MS/MS spectra with the top 40 peak intensities (Supporting Information Table S1) resulted in three marker candidates: NSILTETLHR (residues 659-668, italicized and underlined in Figure 2B; designated as CFTR01), LSLVPDSEQGEAILPR (residues 736-751; designated as CFTR02), and ISVISTGPTLQAR (residues 752-764; designated as CFTR03). Qualification of these three candidate peptides was performed using additional criteria of elution time under the set reversed-phase LC conditions and qualitative overall specificity of MRM transitions. CFTR01 was chosen as the surrogate marker peptide of choice for CFTR quantitation. However, the serine residue (Ser-660) on CFTR01 was reported to be a phosphorylation site.26 Thus, further validation of this peptide marker was conducted using comparative quantitation of CFTR01 in cell lysates that were treated with and without CIP alkaline phosphatase. As shown in Table S2 in the Supporting Information, there was no effect observed for the phosphatase treatment, suggesting that the serine residue did not carry a phosphate group for CFTR samples using the current sample preparation protocols. Routine inclusion of the phosphatase treatment in CFTR sample preparation protocols, however, should make them universally applicable to all samples independent of the phosphorylation state. The purpose for the assay development was to quantify relative changes in CFTR in cultured cells upon drug treatment. In this particular system, one signature peptide was sufficient for the purpose of relative quantitation. One gas-phase transition together with the use of a stable isotope reference peptide that coeluted with the native marker peptide provided satisfactory specificity. In addition, at elution of CFTR01 there were no appreciable interfering peaks. Use of the highest responding MRM transition for CFTR01 measurements allows the ultimate quantitation limit. Use of the CFTR01 peptide for the protein quantitation, instead of the full-length protein, also allowed easier and more robust sample preparation. As shown on lane 10 in Figure 1, due to the hydrophobicity of the transmembrane protein, CFTR aggregated upon storage and showed smear for Western analysis, making the quantitation very complicated and not always reproducible. In contrast, LC-MRM-MS/MS quantitation of CFTR was based on peptides of the protein; the peptides are better analytes than their precursor protein. For relative quantitation, CFTR aggregation was less of a problem; reproducible digestions of CFTR samples were achievable as shown in Supporting Information Table S3. Reference Marker Peptide CFTR01-18O4. With the use of stable isotope labeled peptides as internal standards, the method (25) Fusaro, V. A.; Mani, D. R.; Mesirov, J. P.; Carr, S. A. Nat. Biotechnol. 2009, 27, 190–198. (26) Neville, D. C.; Rozanas, C. R.; Price, E. M.; Gruis, D. B.; Verkman, A. S.; Townsend, R. R. Protein Sci. 1997, 6, 2436–2445.

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

339

of stable isotope dilution allows accurate and precise peptide quantitation by MRM-MS/MS.14,15,17 Stable isotope 18O-labeled peptides have been extensively used for quantitative peptide, protein, and proteome analyses; they are commonly generated by enzymatic catalysis.27-31 A detailed study on slow exchange of peptidyl carboxylate oxygen with water oxygen under weak acidic conditions31 has recently been reported,32 and the labeled peptides with this method have also been proposed for quantitative MS.31,32 Enzymatic 18O-labeling of peptides is the most commonly used method and is efficient, clean, and cost-effective.27-31 Acid catalysis allows 18O-labeling of all peptidyl carboxylates,31,32 which can result in larger mass differences between native and labeled peptides. This mass difference can be beneficial for improving mass separation of differentially labeled, doubly charged precursor ions during MRM analysis on low-resolving triplequadrupole mass spectrometers. It also enhances the flexibility for designing MRM transitions. However, it should be noted that 18 O-labeled reference marker peptides prepared by acid catalysis typically have lower isotope purity than those prepared by enzymatic catalysis, due to the increased number of 18O atoms. The synthetic marker peptide CFTR01 was labeled with water 18 O under acid catalysis, and the reaction progress was monitored with LC-MS to ensure equilibrium oxygen exchange and minimal deamidation byproduct. The reaction was slow, but all carboxylate groups could be labeled (Figure 3, parts A and B). The exchange reaction completed after 85 h when the monoisotopic peak at m/z 592.28 for the doubly charged CFTR01 disappeared completely. The isotope distribution of the 18O-labeled CFTR01 product was deconvoluted into four components (Supporting Information Figure S3), with an assumption that the mass spectrometer authentically analyzed the isotope distribution of peptides. The dominant contributor of the major product peak (doubly charged) at m/z 596.42 was attributed to CFTR01 ions whose four 16O atoms were replaced by 18O (CFTR01-18O4 carried two 18O for the sidechain carboxylate of the glutamic acid residue and two 18O for the C-terminal carboxylate group. This resulted in a total mass increase of 8 Da, or 4 m/z for doubly charged ions). The total amount of this desired labeling product was calculated to be 83%. The most significant minor component at m/z 595.42 was attributed to a byproduct (CFTR01-18O3), due to the residual 16 O in 18O-water used for preparing the labeled peptide. This impurity is about 11%. The larger number of carboxylate groups on the labeled peptide the more significant source of impurity was in the desired per-18O-labeled peptide product. Two minor components were considered, in order to obtain the best mathematical isotope deconvolution of the labeled product. One had the monoisotope at m/z 595.92, and the other had the monoisotope at m/z 597.92. These mass-to-charge ratios were consistent with the likely occurrence of byproduct, which were caused by deamidation of the N-terminal asparagine during (27) Fenselau, C.; Yao, X. J. Proteome Res. 2009, 8, 2140–2143. (28) Yao, X.; Freas, A.; Ramirez, J.; Demirev, P. A.; Fenselau, C. Anal. Chem. 2001, 73, 2836–2842. (29) Yao, X.; Afonso, C.; Fenselau, C. J. Proteome Res. 2003, 2, 147–152. (30) Stewart, I. I.; Thomson, T.; Figeys, D. Rapid Commun. Mass Spectrom. 2001, 15, 2456–2465. (31) Desiderio, D. M.; Kai, M. Biomed. Mass Spectrom. 1983, 10, 471–479. (32) Niles, R.; Witkowska, H. E.; Allen, S.; Hall, S. C.; Fisher, S. J.; Hardt, M. Anal. Chem. 2009, 81, 2804–2809.

340

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

Figure 3. (A) MS spectrum of CFTR01. (B) MS spectrum of 18Olabeling reaction of CFTR01 after 85 h. The major product is CFTR0118 O4 [NSILTE(218O)TLHR(218O)]. (C) MS spectrum of products of CFTR01-18O4 back-exchange reaction in aqueous solution in the presence of trypsin. The major product is CFTR01-18O2 [NSILTE(218O)TLHR].

the acid-catalysis exchange (Supporting Information Figure S3). In general, due to the long time required for producing 18O-labeled peptides by acid catalysis, the deamidation side reaction and its effect on using the resulting labeled peptides as MS quantitation standards need to be considered during experimental design. CFTR01-18O4 has two types of 18O atoms: one on the side chain (two oxygen atoms on the E side chain) and the other at the C-terminal (two oxygen atoms on the C-terminus). The latter is labile to protease-catalyzed exchange reaction with the water oxygen.27,29 This fact can be used to design experiments for monitoring the activity of proteases that are used for protein digestion, which produces targeted marker peptides for MRM quantitation. When CFTR01-18O4 is added during digestion of CFTR-containing protein samples, the two C-terminal 18O atoms can be efficiently back-exchanged with the water oxygen by proteases in the digestion solution. This enzymatic backexchange has no effect on the two 18O atoms on the glutamic acid side chain, resulting in CFTR01-18O2 as shown in Figure 3C. This product can serve two purposes: a reagent for validating the protease activity (to be further discussed in the next section) and an additional internal standard for the CFTR01 quantification, if desired. In addition, use of the CFTR01 peptide with stable isotopes as references provides a check on the validity of an assigned peak in an MRM-MS/MS ion chromatogram for the CFTR native peptide. Peptides with the same sequence but differential numbers (33) Zhang, R.; Sioma, C. S.; Wang, S.; Regnier, F. E. Anal. Chem. 2001, 73, 5142–5149.

Scheme 1. Overview of LC-MRM-MS/MS-Based Quantitation of CFTR

of C, O, and N stable isotopes are known to have minimal differential migration on reversed-phase LC.33 Quantitation of Overexpressed and Endogenous CFTR Using LC-MRM-MS/MS and 18O-Labeled Reference Peptides. CFTR-containing protein samples were first digested with proteases, and the resulting peptide mixtures were measured by LC-MRM-MS/MS for quantifying the marker peptide CFTR01 (Scheme 1). MRM transitions for CFTR01, CFTR01-18O2, and CFTR01-18O4 were m/z 592.3/869.5, m/z 594.3/873.5, and m/z 596.3/877.5, respectively; CFTR01-18O4 was used as the internal standard. As discussed in the above, CFTR01-18O4 can also serve for checking the trypsin activity. To demonstrate this point, as shown in Scheme 1, one portion of CFTR01-18O4 was added right before the addition of trypsin. After the trypsin digestion, this portion of CFTR01-18O4 was converted into CFTR01-18O2 (see Figure 3C). Right before the LC-MS/MS measurements, another portion of CFTR01-18O4 was added as another internal standard for quantitation. As shown in Supporting Information Figure S4, signals obtained for CFTR01-18O2 (broken line in Supporting Information Figure S4A) is very similar to that for CFTR01-18O4 (continued line in Supporting Information Figure S4A), confirming efficient trypsin activity. In principle, both of the labeled CFTR01 can be used as internal standards for the peptide quantitation. Ion chromatograms in Figure 4 had no other major peaks, indicating that the LC-MRM-MS/MS method was highly specific to the native (Figure 4B) and labeled (Figure 4C) CFTR01 peptides. The solid line in Figure 4A was for CFTR01 produced from the digestion of 10 µL of Triton X-100 lysate of BHK cells (equivalent to 0.44 × 106 cells, which was about the limit of quantitation of the current method as shown in Figure 4B). On the basis of a standard calibration curve (Figure 5), the concentration of CFTR01 and thus the parent protein in this sample was calculated. Use of the calibration allowed more accurate CFTR01 quantitation. Although CFTR01-18O4 was expected to have the same ionization efficiency, its signal produced from MRM measurements needed correction, due to the facts that CFTR01-18O4 is not 100% atom pure and the precursor selection for MRM had limited resolution. A standard calibration curve was used to correct the MRM response difference between the native CFTR01 in samples and the

Figure 4. (A) Zoomed ion chromatograms for CFTR01 quantitation using CFTR01-18O4 as an internal standard: ( · · · ) CFTR01-18O4 (596.3/877.5); (s) CFTR01 (592.3/869.5). (B) Ion chromatogram for CFTR01. (C) Ion chromatogram for CFTR01-18O4.

CFTR01-18O4 reference peptide. A series of standard CFTR01 solutions (0, 0.25, 0.5, 1.0, 2.5, and 5.0 fmol/µL, calculated based on amino acid analysis of a standard sample of synthetic CFTR01 peptide) with constant concentration of CFTR01-18O4 (1.0 fmol/µL) were made in aqueous solution with 10% methanol and 0.1% TFA. From the LC-MRM-MS/ MS chromatograms (Supporting Information Figure S5), the ratios of the peak area for CFTR01 to the peak area for CFTR0118 O4 were plotted against the ratios of CFTR01 concentration to CFTR01-18O4 concentration (Figure 5), and the intercept was forced to be zero. A very good linear regression was obtained with a slope of 1.43 (±0.05), which was used to calculate the CFTR amounts in testing samples. In addition, the reproducibility of protein digestion (sample preparation) Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

341

from both cell surface membrane and organelles, whereas CFTR in the biotinylated sample is only from apical plasma membrane. The other possibility is the relatively low efficiency for the biotinylation reaction, due to potentially incomplete surface biotinylation of the membrane CFTR, which has only three theoretical outer membrane amine groups. The efficiency for the surface biotinylation has yet to be examined. HT29 cells express CFTR endogenously, which can also be quantified with help of immunoprecipitation. It is not surprising that these cells have much lower level of CFTR compared to BHK cells.

Figure 5. Standard calibration curve for MRM response ratio vs concentration ratio for CFTR01 to CFTR01-18O4: CFTR01 concentration, 0, 0.25, 0.5, 1.0, 2.5, and 5.0 fmol/µL; CFTR01-18O4 concentration, 1.0 fmol/µL; solvent, aqueous solution with 10% methanol and 0.1% TFA. Table 1. Quantitation of CFTR-Containing Samples CFTR per × 106 cells cells/fraction BHK/Triton X-100 BHK/biotinylation BHK/digitonin (with IP) HT29/Triton X-100 (with IP) a

a

IP antibody

(fmol)

(ng)

M3A7 MAB3484 M3A7

22 ± 1 2.8 ± 0.5 0.41 ± 0.04 0.62 ± 0.06 0.11 ± 0.01

3.7 ± 0.2 0.48 ± 0.08 0.070 ± 0.007 0.11 ± 0.10 0.019 ± 0.002

MAB3484

0.14 ± 0.06

0.024 ± 0.010

Note: IP, immunoprecipitation.

and LC-MRM-MS/MS quantification were examined by replicated measurements from two analysts. Results indicated the good intra- and interanalyst reproducibility for the method (Supporting Information Table S3). All of the measurements were well above estimated LOD and LOQ for the pure CFTR01 peptide (Supporting Information Figure S6). Table 1 summarized quantitation data for different CFTR samples. Each sample was repeated for at least three different preparations. The resulting average and standard deviation are reported in Table 1. For overexpressed CFTR in BHK cells, the CFTR estimated in the Triton X-100 lysate was about 2-fold higher than that in the sample prepared by surface biotinylation. One possible explanation is that CFTR in Triton X-100 lysate comes

342

Analytical Chemistry, Vol. 82, No. 1, January 1, 2010

CONCLUSION The first MS-based assay is demonstrated for quantifying endogenous and overexpressed CFTR in various cellular locations including apical plasma membrane. Enabled by the use of LC-MRM-MS/MS and stable isotope dilution, the high sensitivity, specificity, precision, and reproducibility of the method show clear feasibility for quantifying endogenous apical plasma membrane CFTR in various cells for CF therapeutics development. Further optimization of the method is ongoing for improving the limit of quantitation and throughput. ACKNOWLEDGMENT The authors thank Dr. L. J. DeLucas at University of Alabama at Birmingham, Dr. J. R. Riordan at University of North Carolina, and Dr. J. He at Accelagen for purified CFTR samples and Dr. T. Smith at University of Connecticut for culturing HT29 and BHK cells. We thank Dr. D. R. Wetmore, Dr. E. Joseloff, and Dr. H. Barazi at Cystic Fibrosis Foundation, Dr. B. A. Stanton at Dartmouth Medical School, Dr. C. Bear and Dr. J. Forman-Kay at The Hospital for Sick Children (Canada), and Dr. N. McCarty at Emory University for fruitful scientific discussions and advice. B. A. Coutermarsh at Dartmouth Medical School is acknowledged for producing CFTR samples for preliminary studies, and P. A. C. Diego at University of Connecticut is acknowledged for helping with the reproducibility study of the MRM method as the second analyst to prepare digests of CFTR samples. This work was supported by the Cystic Fibrosis Foundation (YAO07XX0), IGR06-002-01 from the American Cancer Society, and the University of Connecticut. This work was in part presented at the 57th ASMS Conference on Mass Spectrometry and Allied Topics in 2009. SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review November 13, 2009. AC902028F

September

8,

2009.

Accepted