Anal. Chem. 2008, 80, 7169–7173
Analysis of Major Protein-Protein and Protein-Metal Complexes of Erythrocytes Directly from Cell Lysate Utilizing Capillary Electrophoresis Mass Spectrometry An Nguyen and Mehdi Moini* Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712 The separation and detection of the major protein-protein and protein-metal complexes of erythrocytes directly from cell lysate under native conditions has been accomplished for the first time using capillary electrophoresis electrospray ionization-mass spectrometry (CE/ESIMS). All three major protein-protein and protein-metal complexes in human red blood cells (RBCs) with a concentration dynamic range of ∼3 orders of magnitude were successfully detected. Intact complexes detected in lysed RBCs included carbonic anhydrase II (CAII-Zn at ∼0.8 amol/cell) complexed with its zinc cofactor, carbonic anhydrase I (CAI-Zn at ∼7 amol/cell) complexed with its zinc cofactor, and hemoglobin A (Hb-tetramer at ∼450 amol/cell)sa tetramer formed by two r-β-subunits and four heme groups. The average molecular weights measured for these complexes were consistent with their theoretical values within 0.01% mass accuracy. The use of Polybrene as a self-coating reagent in conjunction with ammonium acetate at pH ∼7.4, narrow capillary for high separation efficiency, and forward polarity CE to avoid acid production at the tip of the capillary were overriding experimental factors for successful analysis of protein complexes. Diluting the lysed blood sample in ammonium acetate for a minimum of 6 h before injecting the sample into the CE was essential for obtaining the mass accuracy consistent with their theoretical average molecular weights. At physiological pH, the mass spectrum of the electrophoretic peak of Hb-tetramer included a small amount of the monomers and Hb-dimer. The migration time and peak profile of these species were almost identical to that of the tetramer, indicating that they are formed from decomposition of the Hb-tetramer during the ESI process. A separate electrophoretic peak for the Hb-dimer was only detected when the pH of the BGE was lowered from 7.4 to ∼6.6. Running CE in forward polarity mode was essential for detection of the intact Hb-tetramer as well as CAI-Zn and CAII-Zn complexes. Under forward polarity mode, CE outlet/ESI shared electrode acts as the cathode of the CE circuit and the anode (positive voltage for positive ions) of the ESI circuit, thereby maintaining approximately neutral pH at the CE outlet/ESI electrode. In addition, under forward polarity mode, CAII-Zn and CAI-Zn migrated ahead of Hb-tetramer, avoiding being 10.1021/ac801158q CCC: $40.75 2008 American Chemical Society Published on Web 08/19/2008
masked by 562× and 64×, respectively, molar excess of Hb-tetramer. Recent advances in capillary electrophoresis-mass spectrometry (MS) interfacing and capillary electrophoresis (CE) capillary coating for intact protein analysis using self-coating background electrolytes (BGE) have significantly simplified CE-MS analysis of complex protein mixtures.1,2 Analysis of the protein contents of cells (proteomics) is an important task because the genome itself does not explain many cellular functions and processes. Proteins have specialized functions in the cellular environment and may noncovalently aggregate with DNA, RNA, cofactors, ligands, and other proteins to produce protein complexes. Protein complexes found inside and outside of biological cells take part in many biochemical pathways and perform many different functions. The effects of drugs and other ligands on a particular protein complex can also be investigated, which can become a powerful tool for the biotechnology industry. Mass spectrometry analysis of protein complexes in cells is essential for their identification since the formation of the complex is largely affected by pH and other environmental factors. To date, MS analysis of a mixture of protein complexes is mostly achieved by offline separation and purification of protein complexes followed by MS analysis using infusion electrospray ionization (ESI).3-10 By utilizing this technique, several types of protein complexes have been analyzed including the following: protein-DNA,3 protein-ligand,11 and protein-protein complexes.4-6,12 The main * To whom correspondence should be addressed. Email: mmoini@ mail.utexas.edu. (1) Moini, M. Anal. Chem. 2007, 79, 4241–4246. (2) Garza, S.; Chang, S.; Moini, M. J. Chromatogr., A 2007, 1159, 14–21. (3) Veenstra, T. D. Biochem. Biophys. Res. Commun. 1999, 257, 1–5. (4) Wang, Y.; Schubert, M.; Ingendoh, A.; Franzen, J. Rapid Commun. Mass Spectrom. 2000, 14, 12–17. (5) Light-Wahl, K. J.; Schwartz, B. L.; Smith, R. D. J. Am. Chem. Soc. 1994, 116, 5271–5278. (6) Boys, B. L.; Konermann, L. J. Am. Soc. Mass Spectrom. 2007, 18, 8–16. (7) Loo, J. A. Int. J. Mass Spectrom. 2000, 200, 175–186. (8) Loo, J. A. Mass Spectrom. Rev. 1997, 16, 1–23. (9) Smith, R. D.; Bruce, J. E.; Wu, Q.; Lei, Q. P. Chem. Soc. Rev. 1997, 26, 191–202. (10) Heck, A. J. R.; van den Huevel, R. H. H. Mass Spectrom. Rev. 2004, 23, 368–389. (11) Brenner-Weiss, G.; Kirschhofer, F.; Kuhl, B.; Nusser, M.; Obst, U. J. Chromatogr., A 2003, 1009, 147–153. (12) Martinovic, S.; Berger, S. J.; Pasa-Tolic, L.; Smith, R. D. Anal. Chem. 2000, 72, 5356–5360.
Analytical Chemistry, Vol. 80, No. 18, September 15, 2008
7169
disadvantage of infusion ESI-MS for the analysis of protein complexes is that unbound analytes and salts present in the solution can associate with the protein complexes during the ESI process, resulting in higher average MW for the protein complexes. Moreover, separation and sample cleanup is labor intensive and time-consuming. By separating unbound salts and analytes from the complexes under native conditions, online CE-MS allows faster and more accurate mass measurement of protein complexes directly from cell lysates. CE has been utilized in one instance to detect an intact protein complex;11 however, no separation was accomplished since only one complex was detected. Capillary isoelectric focusing has also been used to separate a mixture of two commercial protein complexes.12 As a proof of the concept, we have chosen human red blood cell lysate (RBCs) to demonstrate the utility of CE/ESI-MS for the analysis of mixtures of protein-protein and protein-metal complexes in a biological sample directly from cell lysate. RBCs are an excellent system for studying intact complexes. Not only are RBCs easy to obtain but they also contain both protein-protein and protein-metal complexes at relatively high concentration levels. Hemoglobin (Hb-tetramer, average MW 64 446) is a protein complex in RBCs that exists as a tetramer consisting of four noncovalently bonded protein subunits: two R-chains (average MW 15 126) and two β-chains (average MW 15 865)13 with each chain attached to a heme group (MW 616). Carbonic anhydrase I (CAI, average MW 28 780)14 and carbonic anhydrase II (CAII, average MW 29 156)15 are two other proteins in RBCs that are each complexed with one zinc (Zn, MW 65) cofactor. This Zn cofactor is very tightly bounded to both CAI (CAI-Zn, MW 28 845) and CAII (CAII-Zn, MW 29 221) at pH ∼7.16 Hb-tetramer, CAI-Zn, and CAII-Zn are the three most abundant protein complexes found in RBCs with quantities of ∼450, ∼7, and ∼0.8 amol/cell, respectively.16 In the past, we have analyzed hemoglobin and carbonic anhydrase (I and II) in intact and lysed RBCs under denatured conditions using CE/ESI-MS.17,18 Other researchers have also analyzed Hb-tetramer under native conditions after offline sample cleanups,5,6 but no online separation and detection of the three major intact protein-protein and protein-metal complexes of lysed RBCs has been reported. Capillary electrophoresis is a powerful separation technique that can be used for the analysis of a variety of compounds from amino acids to intact cells. Two important advantages of CE that are essential for the analysis of protein complexes are as follows: (1) Separation of complexed analytes from analytes not associated with the complex that could otherwise be attached to the complex during electrospray ionization process. CE-MS, therefore, provides a more accurate representation of the constituents of the complex without the use of hydrophobic media that can disintegrate protein complexes. (2) Compatibility with a variety of background electrolytes at a wide pH range, which allows separation of intact protein complexes under native condition. In (13) Hofstadler, S. A.; Severs, J. C.; Smith, R. D.; Swanek, F. D.; Ewing, A. G. Rapid Commun. Mass Spectrom. 1996, 10, 919. (14) Andersson, B.; Nayman, P. O.; Strid, L. Biochem. Biophy. Res. Commun. 1972, 48, 670–677. (15) Henderson, L. E.; Henriksson, D.; Nyman, P. O. Biochem. Biophy. Res. Commun. 1973, 52, 1388–1394. (16) Lindskog, S. Pharmacol. Ther. 1997, 74, 1–20. (17) Cao, P.; Moini, M. J. Am. Soc. Mass Spectrom. 1999, 10, 184–186. (18) Moini, M.; Demars, S. M.; Huang, H. Anal. Chem. 2002, 74, 3772–3776.
7170
Analytical Chemistry, Vol. 80, No. 18, September 15, 2008
this study, we have applied our recently developed techniques for simplifying CE-MS, the porous tip design,1 and the use of bare fused-silica capillary for protein analysis utilizing self-coating background electrolyte,2 to the analysis of intact protein-protein and protein-metal complexes in lysed RBC. EXPERIMENTAL SECTION Sample Preparation. A ∼100-cm-long, 30-µm-i.d., and 150-µmo.d. underivatized fused-silica capillary (Polymicro Technologies, Phoenix, AZ) was used in this experiment. For faster analysis time, pressure-assisted CE-MS was used for CE polarity and pH studies.19 The CE/ESI-MS interface used was the sheathless porous tip design made by etching ∼1.5 in. of the capillary outlet with HF until it was porous.1 The background electrolytes were made by titrating a 0.01% acetic acid solution with a 0.01% ammonium hydroxide solution until desired pHs were reached. At pH of 7.4 the concentration of the ammonium acetate (NH4OAc) buffer was ∼1.7 mM. Polybrene (PB) was added to these buffers as the self-coating reagent2 with a final concentration of PB at ∼0.1%. Unless otherwise mentioned, all solvents were Optima grade (Thermo-Fisher). The fresh blood samples were obtained from a healthy adult and were prepared and lysed according to a previously reported procedure.18 Briefly, 5 µL of blood was diluted to 50 µL using saline solution (0.85%, Sigma) and centrifuged. The supernatant was discarded to remove the plasma, and this process was repeated one more time. The pellet (RBCs) was then diluted 50× and 100× using the NH4OAc buffers containing no PB, which lysed the RBCs. Samples were vortexed for a few seconds to evenly distribute the lysed blood. At 5 nL, the 100× diluted RBCs solution contained ∼129 fmol of HbA, 2 fmol of CAI-Zn, and 229 amol of CAII-Zn.18 Unless otherwise mentioned, the age of the blood is defined as the time between lysing the RBCs and the start of each CE-MS analysis. Four aliquots of whole blood, which were kept in refrigerator (4 °C) for different periods of time, were used to investigate the presence of Hb-dimer in blood, as well as for the study of the reproducibility of the CE-MS analysis. For these experiments, an aliquot of each blood sample was removed from refrigerator, lysed, and analyzed almost immediately by CE--MS. Equipment and Conditions. A P/ACE System MDQ CE instrument (Beckman-Coulter, Fullerton, CA) was used in conjunction with a Q-TOF Premier MS (Waters, Milford, MA) with an m/z range of 8000 on the front quadrupole. Data were analyzed using MassLynx software version 4.1 (Waters). The mass spectra of the intact protein complexes were deconvoluted using Maxent1 software (Waters). The mass spectrometer was optimized and calibrated with sodium trifluoroacetate20 and scanned in the m/z range of 1000-6000. The CE inlet electrode was maintained at +30 (forward polarity mode) or -30 kV (reverse polarity mode). The CE outlet/electrospray electrode voltage was ∼1.5 kV. A sampling cone temperature of 60 °C was used to prevent the decomposition of the protein complexes while maintaining desolvation under ESI. The quadrupole MS profile was set at mass 1200 for 40% dwell time and 20% ramp time, at mass 3000 for 40% dwell time and 0% (19) Cao, P.; Moini, M. Electrophoresis. 1998, 19, 2200–2206. (20) Moini, M.; Jones, B. L.; Rogers, R. M.; Jiang, L. J. Am. Soc. Mass Spectrom. 1998, 9, 977–980.
Figure 1. (A) Base peak electropherogram of blood lysed and diluted 100× in 1.7 mM ammonium acetate (pH 7.4) and kept in the solution for 4 h before CE/ESI-MS analysis with 5 nL of sample injected. (B-D) Corresponding mass spectra of the peaks of panel A marked with CAII-Zn, CAI-Zn, and Hb. The Hb-dimer present in (D) is found to have 2 hemes attached (MW 32 322).
ramp time. Before each separation, the capillary was washed consecutively with 1 M NaOH for 2 min, water for 2 min, methanol for 2 min, and the BGE for 2 min. The washing procedure is essential for detection of low-level protein complexes such as CAIIZn, which can easily be suppressed by up to 562× molar excess of leftover Hb in the capillary. RESULTS AND DISCUSSION Figure 1 shows the electropherogram (panel A) of a lysed blood kept in ammonium acetate for 4 h before analysis, and the corresponding mass spectra (panels B-D) of the CAII-Zn, CAIZn, and Hb-tetramer protein complexes of panel A, respectively. The order of migration, as shown in Figure 1, is CAI-Zn, CAII-Zn, and Hb-tetramer. This order of migration is different from our previous CE-MS analysis of lysed RBCs18 because of the major experimental differences between our older and present CE-MS analyses. In our older publication, RBCs were lysed under denaturing conditions, which dissociated all protein complexes, and were analyzed by CE-MS using denaturing BGE (acidic buffer with pH of ∼3.5). In the present study, protein complexes of RBCs were lysed under native conditions to maintain protein complexes in the solution. In addition, the present study was performed under forward polarity mode whereas the older experiment was performed under reverse polarity mode. For this experiment, low injection volume (∼5 nL) of the 100× diluted lysed blood resulted in high separation efficiency in which CAII-Zn, CAIZn, and Hb were almost baseline separated. The mass spectrum of the electrophoretic peak of the Hb-tetramer (indicated by Q in panel D) is similar to other published reports of this protein complex.5,6 The mass spectrum of panel D also contained a small amount of Hb-dimer [R-β-(heme)2] as indicated by D10+ and D11+,
Figure 2. (A) Base peak electropherogram of blood lysed and diluted 100× in 1.7 mM ammonium acetate (pH 7.4) and kept in the solution for 6 h before CE/ESI-MS analysis with the injection of 4× more sample (20 nL). (B-D) Corresponding mass spectra of the peaks of panel A marked with CAII-Zn, CAI-Zn, and Hb with their corresponding deconvoluted spectrum (inset). The Hb-dimer present in (D) is found to have 2 hemes attached (MW 32 228).
as well as R and β monomers. Extracted ion electropherograms of the monomers and the dimer were almost identical to that of the electrophoretic peak profile of the Hb-tetramer, indicating that they were most likely from the dissociation of Hb-tetramer at the end of the CE capillary and during the desolvation/ionization process (see below for more detail). No separate electrophoretic peak for Hb-dimer or -monomer was observed, even when injecting 4× more sample (20 nL rather than 5 nL) and using a blood sample left in ammonium acetate for 6 h (Figure 2). These results implied that, at physiological pH, and even under sample quantity that overloaded the capillary, the amount of Hb-dimer in lysed blood was too low to be detected by CE/ESI-MS. This is consistent with the binding constant (K) of the Hb-dimer T Hbtetramer, which is reported to be ∼4 × 105 M-1 for fully oxygenated blood (K increases as less oxygen binds) at pH 7.4.21 Comparison of Figures 1 and 2 revealed that electrophoretic peak resolution of Figure 2 was degraded compared to Figure 1 because 4× more sample overloaded the capillary. However, because the blood sample used in Figure 2 had been kept in ammonium acetate solution for a longer period of time before it was analyzed (6 h rather than 4 h in Figure 1), the mass spectral peak resolution of Figure 2 was enhanced, resulting in more accurate MW for the complexes. As shown in Figure 2, the measured average MWs for CAII-Zn, CAI-Zn, and HB-tetramer using Maxent1 deconvolution were respectively 29 221, 28 845, (21) Valdes, R. J.; Vickers, L. P.; Halvorson, H. R.; Ackers, G. K. Proc. Natl. Acad. Sci. U. S. A. 1978, 75, 5493–5496.
Analytical Chemistry, Vol. 80, No. 18, September 15, 2008
7171
Figure 3. Mass spectra of the Hb peak of blood lysed and diluted in 1.7 mM ammonium acetate (pH 7.4) and kept in the solution for (A) 2 and (B) 8 h before CE/ESI-MS analysis. Insets show the closeup of corresponding mass spectra for the mass range of 3300 to 3700. The Hb-dimer present in (A) and (B) are found to have 2 hemes attached. The Hb-monomers (R, β-chains) are found to have heme attached to ∼75% of the monomers (MW 15 743 for R-chain heme and MW 16 484 for β-chain heme) while ∼25% of the monomers do not have heme attached (MW 15 128 for R-chain and MW 15 867 for β-chain).
and 64 456, which compared very well with their corresponding theoretical values of 29 221, 28 845, and 64 446. The accurate MWs obtained from Figure 2 were in contrast with much higher average MWs obtained from deconvoluted mass spectra (not shown) of Figure 1 (29 343, 29 013, and 64 733 for CAII-Zn, CAI-Zn, and Hb-tetramer, respectively.) As shown in Figure 3, the longer the lysed blood stayed in ammonium acetate solution before CE-MS analysis, the more intense and narrow the higher charge states of the protein complexes became. This was because longer times improved the ion exchange between the nonspecific, nonvolatile salts (mostly sodium) that were associated with the interior of the protein complexes under native conditions and the volatile ammonium ions of the BGE. Eventually (after ∼6 h in dilute ammonium acetate solution), the ion exchange resulted in loosening and partial unraveling of the complexes in such a way that, under the CE electric field, complete removal of all nonspecific salt adducts was achieved for a significant population of protein complexes. As shown in Figures 2 and 3, the complete removal of nonspecific salt adducts was achieved for higher charge-state ions only. These higher charge-state ions represented partially unraveled, but intact protein complexes that, under CE-MS became completely desalted, thereby providing accurate (within 0.01% mass spectrometer mass accuracy specification) average MW for protein complexes. For this reason, Q17+, Q18+, and Q19+ were used to obtain average MW for Hb-tetramer, while charge states 11+, 12+, and 13+ were used for CAI-Zn and CAII-Zn. As shown in Figure 3B, even after 8 h in ammonium acetate, the extent of unraveling of protein complexes in blood was not high enough to dissociate the protein complexes, since no separate electrophoretic peaks of Hb chains or Hb-dimer were observed. Migration reproducibility of the protein complexes was studied by analysis of blood samples that were kept in the refrigerator and stored for 1 day or longer. Figure 4 shows the CE-MS analysis of blood samples that were kept in the refrigerator (4 °C) for 1, 4, 9, and 14 days and then lysed and analyzed by CE-MS. As shown, excellent migration reproducibility was observed for protein complexes with an average (for four experiments) RSD of the migration times of 0.4% for Hb-tetramer. In 7172
Analytical Chemistry, Vol. 80, No. 18, September 15, 2008
Figure 4. Base peak electropherograms (mass range 2200-6000) of blood kept in the refrigerator for (A) 1, (B) 4, (C) 9, and (D) 14 days. Blood was then lysed in 1.7 mM ammonium acetate (pH 7.4), diluted 100×, and immediately analyzed by CE/ESI-MS. Insets show the mass spectra of the Hb peak of the corresponding electropherograms. The peak marked with an asterisk (*) is an unidentified peak.
addition, no significant differences or dissociations of the Hb tetramer were observed even for the 14-day-old, refrigerator-kept blood. Under CE forward polarity mode, the only time that a separate electrophoretic peak was observed for Hb-dimer was when the CE BGE was reduced below 6.8. Figure 5 summarizes these results in which systematic CE-MS analyses of lysed RBCs were performed under three BGE pH conditions. As shown, at pH 7.7 (panel A) only Hb-tetramer was observed, while at pH 6.6 (panel B) Hb-tetramer was ∼80% and Hb-dimer was ∼20%. Once the BGE pH was reduced to 5.6 (panel C) only Hb-dimer was observed. These results are consistent with a previous ESI study of lysed RBC at different pH conditions (see end of next paragraph for further information).6 In addition to the pH, the polarity of the CE separation was also found to be a critical factor in observing Hb tetramer. Hemoglobin tetramer can only be observed in forward polarity (Figure 6, panel A). Under reverse polarity, the mass spectrum of the hemoglobin peak was dominated with Hb dimer (Figure 6, panel B). The reason for complete dissociation of the tetramer in reverse polarity mode and minor fragmentation in forward polarity can be explained by the electrochemical reactions at the CE outlet/ESI electrode. As we have explained before, CE and ESIMS represent two electrical circuits each with two sets of electrodes, CE inlet and outlet electrodes, ESI emitter, and MS inlet electrodes.22-25 The CE/ESI-MS overlays these two separate circuits so that the CE outlet electrode and the ES emitter (22) Moini, M. Capillary Electrophoresis/Electrospray Ionization Mass Spectrometry of Amino Acids, Peptides, and Proteins. In Capillary Electrophoresis of Peptides and Proteins; Strege, M. A., Lagu, A. L., Eds.; Methods in Molecular Biology; Humana Press: Totowa, NJ, 2004; Vol. 276, Chapter 13, pp 253-290. (23) Smith, A. D.; Moini, M. Anal. Chem. 2001, 73, 240–246. (24) Moini, M.; Cao, P.; Bard, A. G. Anal. Chem. 1999, 71, 1658–1661. (25) Moini, M. Anal. Bioanal. Chem. 2002, 373, 466–480.
Figure 5. (A) Base peak electropherogram of blood diluted in 1.7 mM ammonium acetate buffer of pH 7.7 showing only Hb-tetramer peak. (B) Base peak electropherogram of blood diluted in 1.7 mM ammonium actate buffer of pH 6.6 showing separate Hb-dimer from Hb-tetramer peak. (C) Base peak electropherogram of blood diluted in 1.7 mM ammonium acetate buffer of pH 5.6 showing only Hb-dimer (with 2 hemes attached) peak. Inset show the mass spectra of the Hb-dimer (with 2 hemes attached) peak of the corresponding electropherogram.
Figure 6. (A) Mass spectra of the Hb in forward polarity and the corresponding chromatogram of blood diluted in 1.7 mM ammonium acetate buffer at pH 7.4 (inset). (B) Mass spectra of the Hb in reverse polarity showing decomposition of tetramer to monomers and dimers and the corresponding electropherogram of blood diluted in 1.7 mM ammonium acetate buffer at pH 7.4 (inset).
electrode are shared between the two circuits (CE outlet/ESI shared electrode). Therefore, under CE/ESI-MS, at the shared electrode two electrochemical reactions occur simultaneously. Depending on the polarity and magnitude of the voltage of the shared electrode compared with that of the CE inlet and MS inlet electrodes, electrochemical reactions at the shared electrode can be both reductive, both oxidative, or one reductive and the other oxidative. For example, under positive electrospray ionization mode, where +1.5 kV is applied to the shared electrode and 0-100 V is applied to the MS inlet electrode, two possibilities exist: (1)
Under reverse polarity CE, where ∼-30 kV is applied to the CE inlet electrode, the shared electrode is anodic with respect to both CE inlet electrodes and the MS inlet electrode. (2) Under forward polarity CE, where ∼+30 kV is applied to the CE electrode, the shared electrode is cathodic with respect to CE inlet electrode and anodic with respect to MS inlet electrode. At the anode, the major oxidation reaction for aqueous buffer solutions is typically the electrochemical oxidation of water (2H2O S 4H+ + O2 + 4e-). The extent of the electrochemical reactions and therefore the pH change depends on the magnitude of the current that flows through each circuit. Under experimental conditions used here, CE current was ∼1 µA while ESI current was in the nanoampere range. Under reverse polarity mode in which the electrochemical reactions at the shared electrode are anodic for both circuits, the pH of the solution decreases enough to dissociate all the tetramer. The pH at the ESI tip under reverse polarity is estimated to be decrease from pH 7.4 to ∼5, the pH at which Hb-tetramer completely dissociates to dimer.6 However, under forward polarity mode, the shared electrode is only anodic in the ESI circuit, which has a lower current, and its effect is mostly canceled by the cathodic reaction at this electrode in the CE circuit causing minimal dissociation of the tetramer. Since this dissociation (in both cases) is happening close to the outlet of the CE capillary, all dissociation products comigrate and show up as a single electrophoretic peak. CONCLUSION The analysis of intact protein-protein and protein-metal complexes of the lysed RBC was accomplished for the first time using CE/ESI-MS. Intact CAII-Zn, CAI-Zn, and Hb-tetramer complexes were analyzed with no prior sample preparation. Accurate MWs (∼0.01% error) were obtained for all major protein-protein and protein-metal complexes of lysed RBCs. It was observed that more accurate MW was obtained when protein complexes were diluted in ammonium acetate buffer and left in that solution for a few hours. Ion exchange between ammonium ions in the solvent and nonspecific sodium adducts of the protein-protein and protein-metal complexes slightly unraveled the complexes. CE/ESI-MS further enhanced this charge exchange and separated the nonspecific salt adducts from protein complexes, resulting in formation of higher charge-state protein complexes free from nonspecific salt adducts. Deconvolution of these higher charge-state ions provided accurate (∼0.01% error) average MW of intact protein-protein and protein-metal complexes for better identification of the constituents of the complexes. Operation of the CE under forward polarity mode was essential for detection of intact protein-protein and protein-metal complexes. This was because of the electrochemical nature of CE and ESI processes. CE/ESI-MS analysis of lysed blood at pH of ∼7.4 showed no separate peak for Hb-dimer, indicating that hemoglobin in fresh blood is mostly in its tetrameric form. These results demonstrate that CE/ESI-MS is an effective tool for the analysis of the mixture of protein complexes and can accurately provide average MW for unambiguous identification of small (
