Article pubs.acs.org/ac
Targeted Absolute Quantification of Intact Proteins by Reversed Phase Liquid Chromatography-Mass Spectrometry, Charge Reduced Electrospray, and Condensation Particle Counting Kouame Adou,* Murray V. Johnston, and John L. Dykins Department of Chemistry and Biochemistry, University of Delaware, Newark, Delaware 19716, United States ABSTRACT: A novel approach involving the use of reversed phase liquid chromatography-mass spectrometry (RPLC-MS), charge reduced electrospray (CRES), and condensation particle counting (CPC) for the absolute quantification of intact proteins in liquid solutions is introduced. Under analysis conditions optimized for the quantification of select proteins within their predetermined linear ranges, a set of at least five protein standards with molecular weights (MW) spanning the dynamic ranges of both a quadrupole time-of-flight (QTOF) MS and a suitably selected RPLC column is used to generate a calibration curve of CPC detection efficiency (DE) as a function of the square root of MW. Next, the sample of interest is analyzed, and from the MS-generated MW data, the DE of each target protein is determined from the calibration curve. On the basis of MW, DE, and number concentration (molecules/unit volume), absolute quantification is achieved for each protein of interest. Application of this approach to the absolute quantification of cytochrome C (as target compound) in a commercial protein mixture is demonstrated with a deviation of 8%, a coefficient of variation (CV) of 5%, and a quantification limit of 432 fmol. For nontarget components of the mixture (ribonuclease A, holotransferrin, and apomyoglobin), the percent deviation from the stated concentrations and the CV varied from 0.20 to 23 and from 4.1 to 18, respectively. Performance of the method was further assessed by analyzing a laboratory quality control mixture comprising 0.33 μM of cytochrome C. The calculated value was 0.34 (CV: 5.1%). Universal in essence, the new technique holds strong promise for the absolute quantification of select proteins in liquid samples under conditions of good peak resolution and stable baseline.
A
(known) in liquid solutions. The use of chromatography and CPC for analyte quantification was first investigated by the Koropchack group,26−30 but studies by this team did not make use of mass spectrometry and charge reduction. In a more recent study, Hutchinson et al.31 assessed the use of four different aerosol detectors (as universal detectors) with ultra high pressure liquid chromatography, but unlike the technique introduced here, their investigation involved neither protein samples nor mass spectrometry. More than a decade ago, Lewis and coworkers investigated the use of RPLC, CRES, and CPC for the absolute quantification of intact proteins32 in protein mixtures. Although nearly equal number concentrations were expected for equimolar amounts of analyte, the observed peak areas for a mixture of six proteins were not in agreement with the expected ratios. On the basis of these results, the authors disqualified the CPC as a good quantitative detector for RPLC analysis of protein mixtures. While the work described here also involves the use of RPLC, CRES, and CPC, it does heavily rely on MS-generated molecular data and the concept of CPC detection efficiency.33 From a set of model proteins, it is shown that, under CRES conditions, CPC detection efficiency is linearly related to the square root of protein molecular weight and that CPC-generated
lthough recent advances in quantitative proteomics have led to the discovery of a significant number of proteins that may be used to monitor certain diseases, subsequent validation and clinical use of these biomarkers have been significantly dampened by the lack of “universal” quantification procedures.1 The enzyme-linked immunosorbent assay (ELISA)2,3 is the current preferred method for absolute protein quantification.1,4 However, very few ELISA antibodies exist for the discovered proteins, and it does not appear cost-effective to develop more antibodies for the purpose of biomarker validation. In recent research, Appleyard et al.4 introduced a particle-based multiplexed quantification method but this technique also requires the use of specific antibodies. As alternatives to ELISA, mass spectrometry-based quantification techniques are increasingly being developed in quantitative proteomics.5−17 In their vast majority, however, these procedures are based on the use of stable isotopically labeled amino acids,18−23 a strategy that often leads to severely compromised results due to a number of factors including poor sequence coverage7,18,24 and inaccurate quantification.18,25 Because of these limitations, the development of label-free quantification procedures has been strongly encouraged. The purpose of this work was to use reversed phase liquid chromatography (RPLC), mass spectrometry (MS), charge reduced electrospray (CRES), and condensation particle counting (CPC) to identify and quantify intact target proteins © 2012 American Chemical Society
Received: April 9, 2012 Accepted: July 17, 2012 Published: July 17, 2012 6981
dx.doi.org/10.1021/ac300950w | Anal. Chem. 2012, 84, 6981−6985
Analytical Chemistry
Article
number concentrations may be used for the absolute quantification of nearly any protein in liquid solutions. For globular proteins, application of this procedure to the absolute quantification of cytochrome C in a commercial HPLC protein mixture is demonstrated.
■
EXPERIMENTAL SECTION Materials. Protein reference materials (insulin and trypsinogen from bovine pancreas, cytochrome C from bovine heart, alpha- and beta-lactoglobulin from bovine milk, and bovine serum albumin) and a protein powder comprising 1 mg each of cytochrome C, ribonuclease A, holotransferrin, and apomyoglobin were obtained from Sigma-Aldrich (Saint Louis, MO). Trifluoroacetic acid, 1-butanol, and acetonitrile were obtained from Fisher Scientific (Pittsburgh, PA). Protein stock solutions were prepared in deionized water (in-house) with the exception of insulin which was prepared in 0.1% trifluoroacetic acid (TFA). Prior to use, all sample solutions (in 10% acetonitrile) and mobile phases were filtered through a 0.22 μm membrane filter (Millipore Corporation, Billerica, MA) and sonicated under vacuum for at least 10 min. Charge Reduced Electrospray Interface. The CRES interface used in this study has been described in detail elsewhere.33 Briefly, a virgin electrical grade polytetrafluoroethylene (PTFE) bar (6 in. × 4 in. × 2 in., McMaster-CARR, New Brunswick, NJ, part number 8743k47) was machined to house a Micromass Q-TOF Ultima API-US nanoflow source (Micromass, Manchester, UK) and subsequently fastened to an XY positioning slide (McMaster-CARR, New Brunswick, NJ, part number 60935k27). Here, the source was equipped with a 2 in. stainless steel megaflow capillary (125 μm ID × 320 μm OD, Waters, Milford, MA) for microspray operation. A stainless steel cone (2 in. length, 3 mm ID) was connected to a 210Po source (10 mCi, NRD, Grand Island, NY) for in-line sampling and charge reduction. Simultaneous RPLC-MS and RPLC-CRES-CPC Analysis. Mass spectra and number concentration data were simultaneously collected from each RPLC-analyzed solution by positioning the CRES electrospray ionization (ESI) source between the sampling cone of a commercial QTOF mass spectrometer equipped with MaxEnt1 software (Ultima API-US, Micromass, Manchester UK) and that of the CRES interface connected to a condensation particle counter (Model 3775, TSI, Inc., Shoreview MN). The related experimental setup is shown in Figure 1. RPLC was carried out by injecting samples onto a liquid chromatograph system (Models LC-20AD, CTO-20A, CBM20A; Shimadzu, Columbia, MD) equipped with a 10 μL loop and a protein RPLC column (Michrom Bioresources, Auburn, CA; PLRP-S 5 μ 1000 A, 1.0 × 150 mm, P/N # CM5/00711/00). Mobile phases A and B were 10% acetonitrile and 0.1% TFA and 90% acetonitrile and 0.1% TFA, respectively. Elution conditions included an isocratic hold at 15% B for 4 min, three successive increases in B content to 28, 38, and 55% from 4 to 12 min, 12 to 26 min, and 26 to 45 min, respectively, and a hold at 55% B for 5 min followed by a 15 min equilibration step. The entire HPLC effluent, at a rate of 40 μL/min, was directed to the ESI source. The same ESI aerosol was sampled simultaneously by both analytical instruments as described previously. 33 Aerosol sampling into the CPC was performed approximately 5 mm from the axis of the spray tip and 15 mm orthogonally to the aerosol plume. The quadrupole time-of-flight (QTOF) sampling cone and the ESI source were held at potentials of 60 V and 4.8 kV, respectively.
Figure 1. Simultaneous RPLC-MS and RPLC-CPC Analysis.
■
RESULTS AND DISCUSSION RPLC-CRES-CPC Linear Range and Detection Limit. The linear dynamic range of the RPLC-CRES-CPC system was determined using standard solutions of the target analyte (cytochrome C). A stock solution of this protein was prepared in deionized water and used to produce working solutions of 0.04 to 2 μM in 20% ACN and 0.1% TFA. For each solution, an aliquot of 10 μL was injected and analyzed under the conditions described in the Experimental Section. The time-dependent CPC raw data were processed as described before.33 From a plot of molecules detected as a function of cytochrome C concentration, it was found that CPC response was linear from 0.04 to 0.8 μM. The lower limit of quantification was 432 fmol, which compares well with recent peptide-based targeted absolute quantification techniques.12 The linear range represents the range of protein concentrations for which most of the ESI droplets contain a single molecule. Over this range, most particles detected and counted by the CPC represent one protein molecule. Beyond the upper point of linearity, most of the aerosol particles detected by the CPC contain more than one molecule (association of molecules in the liquid phase at high concentrations). Compared to our previous results,33 the observed linear range is smaller. One possible explanation could be based on the spray solvent used here. Under the same flow and source conditions, a solvent system of 30% methanol would usually spray better than a system of 20−45% ACN and 0.1% TFA (elution region of the analyzed proteins) due to signal suppression by TFA. Nonetheless, it was sufficiently wide for the purpose of this study. RPLC-CRES-CPC Calibration. Following the above linearity study, standard mixtures of two or more components including insulin (5.8 kDa), cytochrome C (12.3 kDa), beta-lactoglobulin (18.2 kDa), alpha-lactoglobulin (18.6 kDa), trypsinogen (23.8 kDa), and bovine serum albumin (66 kDa) were prepared and analyzed at the level of 0.35 μM in 20% ACN and 0.1% TFA, in RPLC-CRES-CPC. While this solution concentration was previously determined to be within the linear range of the RPLC-CRES-CPC system for the target analyte, it was also 6982
dx.doi.org/10.1021/ac300950w | Anal. Chem. 2012, 84, 6981−6985
Analytical Chemistry
Article
electrospray aerosol into the transfer line is most likely the greatest contributor to particle loss, since most of the electrospray current flows to the counter electrode. From the linear relationship between DE and molecular weight, the following formulas were established for the absolute quantification of intact globular proteins.
assumed to primarily produce droplets with zero or one molecule of the other proteins based on previous studies. A related sample CPC chromatogram is shown in Figure 2. For each protein, the
C=
m×M V × N × DE
In this formula, C is the sample concentration in μg/mL, m is the CPC-generated molecules count, M is the molar mass (derived from the MS-determined molecular weight) in μg/mol, V is the volume injected in mL, N is the Avogadro number, and DE is the CPC detection efficiency for the analyzed protein. Application to the Targeted Quantification of Cytochrome C in a Protein Mixture. Using the analytical conditions described earlier, the validity and accuracy of the above quantification procedure were assessed by analyzing a commercial protein powder comprising 1 mg each of cytochrome C, ribonuclease A, holotransferrin, and apomyoglobin. From a stock solution of the sample prepared in water, a dilute working solution of 12 μg/mL (0.98 μM cytochrome C) was prepared in 20% ACN and 0.1% TFA and injected following analysis of the molecular weight calibrants. The related CRESCPC and QTOF total ion chromatograms (TIC) are shown in Figure 4a,b, respectively. Each peak in the MS TIC was scanned, and the related mass spectrum was examined. Identification of cytochrome C was based on QTOF molecular weight data. Peak 3 in the TIC (peak at 20.36 min in Figure 4b) of the analyzed sample gave an average spectrum (Figure 4c) exhibiting positive electrospray charge states of +8 (1529 m/z), +7 (1748 m/z), +6 (2039 m/z), +5 (2446 m/z), and +4 (3057 m/z), which is consistent with cytochrome C, a globular protein of approximately 12.23 kDa. The MS identification was confirmed by the elution time and MS spectrum of cytochrome C in a previously analyzed standard solution. From a10 μL aliquot of the sample mixture, an average (n = 2, coefficient of variation (CV) = 5.0) of 2.97 × 107 molecules of cytochrome C were detected by the CPC over the elution time of this compound. Using the above quantification equation, the MS-determined molecular weight of 12.2 kDa, and the corresponding DE, a cytochrome C sample concentration of 11 μg/mL (0.90 μM) was calculated (deviation from the known value by less than 10%). Beside cytochrome C, absolute quantification was also carried out for nontarget proteins in the powder. From the prepared solution (12 μg/mL), the calculated sample concentrations were 9.7, 12, and 9.3 μg/mL for ribonuclease A, apomyoglobin, and holotransferrin, respectively. A summary of these quantification results is shown in Table 1.
Figure 2. RPLC-CRES-CPC chromatogram of a 0.35 μM protein standard mixture comprising insulin (1), cytochrome C (2), trypsinogen (3), alpha lactoglobulin (4), bovine serum albumin (5), and myoglobin (6). Injection volume: 10 μL. Solvent: 20% ACN and 0.1% TFA.
total number of molecules was calculated and the detection efficiency33 of the system was determined. Figure 3 shows the
Figure 3. CRES-CPC detection efficiency (DE) as a function of molecular weight. Insulin (5.8 kDa), cytochrome C (12.2 kDa), betalactoglobulin (18.2 kDa), alpha-lactoglobulin (18.6 kDa), trypsinogen (28.3 kDa), and bovine serum albumin (66 kDa) were used, within their linear ranges at the level of 0.35 μM, to generate the above curve. Injection volume: 10 μL. Error bars are standard deviations of duplicate runs.
■
CONCLUSIONS One benefit of the introduced method resides in its ability to quickly quantify intact proteins, which eliminates the need for preliminary digestion. Moreover, the initial plot of DE as a function of molecular weight may be generated from any set of suitable calibrants and used for the targeted quantification of any globular protein with a molecular weight less or approximately equal to the upper point of calibration. The lower limits of detection and quantification may be significantly reduced by (1) increasing the particles intake of the CPC (for example, utilizing a smaller but robust spray tip) and (2) using a mass spectrometer of the newer generation and/or employing an RPLC column that does not require the use of TFA.
variation of detection efficiency (DE) with respect to molecular weight. The data suggest that for globular proteins, at linear range concentrations, CPC DE is linearly correlated to the square root of the molecular weight. The observed increase of CPC DE with molecular size is consistent with the work of Lewis et al.34 showing that smaller and lighter molecules diffuse more rapidly and are more likely to experience wall loss in the transfer line. In the above setup (Figure 1), however, sampling from the 6983
dx.doi.org/10.1021/ac300950w | Anal. Chem. 2012, 84, 6981−6985
Analytical Chemistry
Article
Figure 4. (a) RPLC-CRES-CPC chromatogram of a commercial protein mixture comprising 1 mg each of ribonuclease A (2), cytochrome C (3), holotransferrin (4), and apomyoglobin (5); peak 1 is the solvent front; (b) related RPLC-QTOF total ion chromatogram; (c) QTOF average mass spectrum of cytochrome C showing the +8 (1529 m/z), +7 (1748 m/z), +6 (2039 m/z), +5 (2446 m/z), and +4 (3057 m/z) charge states.
Table 1. Absolute Quantification Resultsa calculated μM ribonuclease A cytochrome C apomyoglobin holotransferrin a
MW (kDa)
DE
known μM
average (n = 2)
CV
% error
13.7 12.2 16.6 77
5.49 × 10−6 5.15 × 10−6 6.21 × 10−6 1.45 × 10−5
0.88 0.98 0.71 0.16
0.71 0.90 0.71 0.12
4.1 5.0 7.9 17.6
19 8 0.2 23
MW: molecular weight. DE: detection efficiency.
■
For accurate results, however, a few conditions must be satisfied. First, the HPLC conditions must be adjusted to elute the peak(s) of interest at relatively stable baseline. CPC noise increases as baseline rises in gradient elution, which reduces accuracy. Second, the targeted protein(s) must be well resolved from other compounds. Lastly, the analyzed proteins must be within the dynamic ranges of the mass spectrometer and the RPLC column used. When these criteria are met, this novel technique is a robust (compatible with most RPLC mobile phases) and fast procedure for the absolute quantification of known proteins in liquid samples. For example, it may be utilized for the rapid analysis of clinical samples in which the concentrations of specific biomarkers are needed to monitor the state of certain diseases.
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This research was supported by NIH-INBRE Grant Number 5P20RR016472. TSI Inc. provided the Condensation Particle Counter used in this study.
■
REFERENCES
(1) Pan, S.; Aebersold, R.; Chen, R.; Rush, J.; Goodlett, D. R.; McIntosh, M. W.; Zhang, J.; Brentnall, T. A. J. Proteome Res. 2008, 8, 787−797. 6984
dx.doi.org/10.1021/ac300950w | Anal. Chem. 2012, 84, 6981−6985
Analytical Chemistry
Article
(2) Gonzalez, R. M.; Seurynck-Servoss, S. L.; Crowley, S. A.; Brown, M.; Omenn, G. S.; Hayes, D. F.; Zangar, R. C. J. Proteome Res. 2008, 7, 2406−2414. (3) Lequin, R. M. Clin. Chem. 2005, 51, 2415−2418. (4) Appleyard, D. C.; Chapin, S. C.; Doyle, P. S. Anal. Chem. 2010, 83, 193−199. (5) Barnidge, D. R.; Goodmanson, M. K.; Klee, G. G.; Muddiman, D. C. J. Proteome Res. 2004, 3, 644−652. (6) Bouchal, P.; Roumeliotis, T.; Hrstka, R.; Nenutil, R.; Vojtesek, B.; Garbis, S. D. J. Proteome Res. 2008, 8, 362−373. (7) Brun, V.; Dupuis, A.; Adrait, A.; Marcellin, M.; Thomas, D.; Court, M.; Vandenesch, F.; Garin, J. Mol. Cell. Proteomics 2007, 6, 2139−2149. (8) Grant, J. E.; Bradshaw, A. D.; Schwacke, J. H.; Baicu, C. F.; Zile, M. R.; Schey, K. L. J. Proteome Res. 2009, 8, 4252−4263. (9) Hägglund, P.; Bunkenborg, J.; Maeda, K.; Svensson, B. J. Proteome Res. 2008, 7, 5270−5276. (10) Hanke, S.; Besir, H.; Oesterhelt, D.; Mann, M. J. Proteome Res. 2008, 7, 1118−1130. (11) Ji, Q. C.; Rodila, R.; Gage, E. M.; El-Shourbagy, T. A. Anal. Chem. 2003, 75, 7008−7014. (12) Li, N.; Palandra, J.; Nemirovskiy, O. V.; Lai, Y. Anal. Chem. 2009, 81, 2251−2259. (13) Lokaj, K.; Meierjohann, S.; Schütz, C.; Teutschbein, J.; Schartl, M.; Sickmann, A. J. Proteome Res. 2009, 8, 1818−1827. (14) Orlando, R.; Lim, J.-M.; Atwood, J. A.; Angel, P. M.; Fang, M.; Aoki, K.; Alvarez-Manilla, G.; Moremen, K. W.; York, W. S.; Tiemeyer, M.; Pierce, M.; Dalton, S.; Wells, L. J. Proteome Res. 2009, 8, 3816−3823. (15) Pan, S.; Rush, J.; Peskind, E. R.; Galasko, D.; Chung, K.; Quinn, J.; Jankovic, J.; Leverenz, J. B.; Zabetian, C.; Pan, C.; Wang, Y.; Oh, J. H.; Gao, J.; Zhang, J.; Montine, T.; Zhang, J. J. Proteome Res. 2008, 7, 720− 730. (16) Shah, S. J.; Yu, K. H.; Sangar, V.; Parry, S. I.; Blair, I. A. J. Proteome Res. 2009, 8, 2407−2417. (17) Xu, P.; Duong, D. M.; Peng, J. J. Proteome Res. 2009, 8, 3944− 3950. (18) Collier, T. S.; Hawkridge, A. M.; Georgianna, D. R.; Payne, G. A.; Muddiman, D. C. Anal. Chem. 2008, 80, 4994−5001. (19) Ghosh, D.; Yu, H.; Tan, X. F.; Lim, T. K.; Zubaidah, R. M.; Tan, H. T.; Chung, M. C. M.; Lin, Q. J. Proteome Res. 2011, 10, 4373−4387. (20) Keshamouni, V. G.; Jagtap, P.; Michailidis, G.; Strahler, J. R.; Kuick, R.; Reka, A. K.; Papoulias, P.; Krishnapuram, R.; Srirangam, A.; Standiford, T. J.; Andrews, P. C.; Omenn, G. S. J. Proteome Res. 2009, 8, 35−47. (21) Magharious, M.; D’Onofrio, P. M.; Hollander, A.; Zhu, P.; Chen, J.; Koeberle, P. D. J. Proteome Res. 2011, 10, 3344−3362. (22) Molina, H.; Yang, Y.; Ruch, T.; Kim, J.-W.; Mortensen, P.; Otto, T.; Nalli, A.; Tang, Q.-Q.; Lane, M. D.; Chaerkady, R.; Pandey, A. J. Proteome Res. 2008, 8, 48−58. (23) Ralhan, R.; DeSouza, L. V.; Matta, A.; Chandra, T. S.; Ghanny, S.; DattaGupta, S.; Thakar, A.; Chauhan, S. S.; Siu, K. W. M. J. Proteome Res. 2008, 8, 300−309. (24) Patel, V. J.; Thalassinos, K.; Slade, S. E.; Connolly, J. B.; Crombie, A.; Murrell, J. C.; Scrivens, J. H. J. Proteome Res. 2009, 8, 3752−3759. (25) Ow, S. Y.; Salim, M.; Noirel, J.; Evans, C.; Rehman, I.; Wright, P. C. J. Proteome Res. 2009, 8, 5347−5355. (26) Allen, L. B.; Koropchak, J. A.; Szostek, B. Anal. Chem. 1995, 67, 659−666. (27) Guo, W.; Koropchak, J. A.; Yan, C. J. Chromatogr., A 1999, 849, 587−597. (28) Koropchak, J. A.; Heenan, C. L.; Allen, L. B. J. Chromatogr., A 1996, 736, 11−19. (29) Szostek, B.; Koropchak, J. A. Anal. Chem. 1996, 68, 2744−2752. (30) Yang, X.; Koropchak, J. A. J. Microcolumn Sep. 2000, 12, 204−210. (31) Hutchinson, J. P.; Li, J.; Farrell, W.; Groeber, E.; Szucs, R.; Dicinoski, G.; Haddad, P. R. J. Chromatogr., A 2011, 1218, 1646−1655. (32) Lewis, K. C.; Jorgenson, J. W.; Kaufman, S. L.; Skogen, J. W. J. Microcolumn Sep. 1998, 10, 467−471. (33) Adou, K.; Johnston, M. V. Anal. Chem. 2009, 81, 10186−10192.
(34) Lewis, K. C.; Dohmeier, D. M.; Jorgenson, J. W.; Kaufman, S. L.; Zarrin, F.; Dorman, F. D. Anal. Chem. 1994, 66, 2285−2292.
6985
dx.doi.org/10.1021/ac300950w | Anal. Chem. 2012, 84, 6981−6985
