Influence of Instrumental Conditions on the Electrospray Ionization Mass Spectrometry of Peptides/Proteins in Mixtures

The simultaneous detection of several peptides or proteins in mixtures is an integral part of expression proteomics investigations. Strategies that enable the identification and characterization of several proteins or peptides simultaneously from complex mixtures with minimal recourse to sample cleanup or separation stages prior to detection are ideally desired for high-throughput proteomic investigations. The advent of soft ionization mass spectrometric techniques, such as matrix-assisted laser desorption ionization-mass spectrometery (MALDI-MS) and electrospray ionization-mass spectrometery (ESI-MS), has enabled different strategies to be developed based on these techniques.1–4 In particular, the higher mass accuracy, sensitivity, and better reproducibility of analysis possible with ESI- (and nanospray-) MS have allowed its broader application in expression proteomic investigations. It has been employed in both “top-down” (involving analysis of “intact” proteins by tandem mass spectrometry without recourse to enzymatic proteolytic digestions)5,6 and “bottom-up” strategies (involving the detection of peptides from digested proteins)7,8 for proteomic characterizations.

In ESI-MS analysis, several factors influence signal detection, including the analyte concentration, pH and ionic strength of the medium, nature of the analyte, and instrumental parameters.9 For proteins, these factors are known to influence the intensity and charge state distribution of the protein peaks, even when the proteins are analyzed in isolation.9–12 When considering the analysis of proteins or peptides in mixtures, the ionization of the individual components and the presence of other components can affect the detection of any individual peptide or protein.13 Although it has been recognized that several factors contribute to the formation of gas-phase ions in ESI-MS, understanding the behavior of ions for the efficient analysis of proteins or peptides in mixtures is far from complete. Once gas-phase ions are generated in the source, they must still traverse the source–analyzer interface before being detected. This presents an additional set of factors that can influence the protein or peptide signals detected. Skimmer voltages, pressure, and ion transmission optics have all been shown to be influential in the detection of appropriate protein signals.10,14–17 The tuning of instrumental settings in the source and source–analyzer interface is thus likely to influence the detection of protein signals in mixtures.

Figure 1 - Schematic showing instrumental configuration of the ESI-QToF mass spectrometer used in the study. The areas associated with the 14 instrumental settings are numbered in the figure and listed in the table alongside. The regions of spectrometer associated with ion generation, acceleration, transmittance, and detection are also indicated.

Investigations conducted by the author and colleagues18,19 have demonstrated this aspect with a cocktail of five proteins in the mass range of 5–17 kDa using a Micromass (now Waters, Milford, MA) ESI-QToF mass spectrometer. The mixture of five proteins, including insulin (5733 Da), ubiquitin (8565 Da), cytochrome c (12361 Da), lysozyme (14309 Da), and myoglobin (16951 Da), was dissolved in equimolar concentrations and analyzed in acidic conditions in the positive ion mode. The concentration used was low enough to maintain conditions of excess charge and linearity of signal response, so that any change in signal observed is not due to charge limitations or concentration differences but due solely to the nature of the protein and its behavior under a given set of instrumental conditions. Fourteen instrumental settings (Figure 1) that influence ion generation, acceleration, transmission, and detection within the mass spectrometer were studied. Over 400 combinations of the instrumental settings were analyzed.

Figure 2 - Pseudo 3-D plot of the variable search space in the first three principal components (PC1,2,3) showing the distribution of the trials where each trial indicates a set of instrumental conditions. Trials in which uniform detection of all five proteins (A) and the preferential detection of cytochrome c (B), ubiquitin (C), lysozyme (D), myoglobin (E), and insulin (F) in the mixture are circled, and their corresponding mass spectrum is shown. The difference in the spectra between the different cases can be clearly seen.

The responses of the individual proteins when analyzed in isolation differed from their responses when analyzed in the mixture, even for a given instrumental setting.18 The instrumental settings had a considerable effect on the detection of the protein peaks, so much so that under certain conditions selective detection of the individual proteins in complete exclusion of the others in the mixture was possible.19 The protein signals were influenced to varying degrees under different instrumental settings. An analysis of the instrumental settings in relation to the responses shows that it is possible to identify areas in the search space in which the conditions permit selective detection of one or more proteins in the mixture and in which all five proteins could be evenly detected (Figure 2).

The problem of identifying unique values of the settings for detecting the proteins preferentially or evenly is complicated by its epistatic nature (i.e., the effect of a given setting can depend on the values of the others). However, it should be possible to optimize the settings for a desired effect, as has been demonstrated by the author and colleagues for the case of uniform detection of all five proteins in the mixture.18 Given the number of influencing factors, their levels, and the resulting combinations to experiment, heuristic methods will need to be employed,20 which seek solutions that approach an optimum but cannot be guaranteed to find it, such as evolutionary algorithms (EAs), in navigating the search spaces to find regions of optimality for a desired effect.

The utility of such computational methods is demonstrated by the fact that for the uniform detection of all five proteins in the mixture, the use of an EA in the search on a search space that was a nominal 1014 experiments resulted in convergence toward optimality, even before 500 of them were evaluated.18 The added advantage of such computational strategies is that in addition to being exploratory, they can be explanatory in that they can sometimes provide an explanation for the optimal behavior. In this particular experiment, it was noted that much of the desired result of the uniform detection of all five proteins arose from maintaining a defined skimmer cone potential between the first and second skimmer (parameters 7 and 8 in Figure 1) in the instrumental setup.18

In addition to the instrumental settings, differences were also observed in the protein profiles from bacterial cell extracts when the solvent conditions were varied. Clearly, the experimental conditions employed must be given due consideration for the efficient detection of proteins or peptides in mixtures. Moreover, the ability to vary the instrumental settings rapidly means that we may hope to be able to detect and analyze imperfectly separated proteins/peptides on-line and in real-time for proteomic applications. The tuning of instrumental settings in the source and the source–analyzer interface is thus likely to provide an important tool for developing strategies for the efficient analysis of proteins or peptides in mixtures, eventually enabling wider proteome coverage than is possible with current technology.

References

  1. Smith, R.D. Trends in mass spectrometry instrumentation for proteomics. Trends Biotechnol. 2002, 20, S3–7.
  2. Lee, S.W.; Berger, S.J.; Martinovic, S.; Pasa-Tolic, L.; Anderson, G.A.; Shen, Y.F.; Zhao, R.; Smith, R.D. Direct mass spectrometric analysis of intact proteins of the yeast large ribosomal subunit using capillary LC/FTICR. Proceedings of the National Academy of Sciences of the United States of America2002, 99, 5942–7.
  3. Aebersold, R.; Mann, M. Mass spectrometry-based proteomics. Nature2003, 422, 198–207.
  4. Hayter, J.R.; Robertson, D.H.L.; Gaskell, S.J.; Beynon, R.J. Proteome analysis of intact proteins in complex mixtures. Molec. Cell. Proteom. 2003, 2, 85–95.
  5. Reid, G.E.; McLuckey, S.A. “Top down” protein characterization via tandem mass spectrometry. J. Mass Spec. 2002, 37, 663–75.
  6. Stephenson, J.L.; McLuckey, S.A.; Reid, G.E.; Wells, J.M.; Bundy, J.L. Ion/ion chemistry as a top-down approach for protein analysis. Curr. Opin. Biotechnol. 2002, 13, 57–64.
  7. Wysocki, V.H.; Resing, K.A.; Zhang, Q.; Cheng, G. Mass spectrometry of peptides and proteins. Methods2005, 35, 211–22.
  8. Bogdanov, B.; Smith, R.D. Proteomics by FTICR mass spectrometry: top down and bottom up. Mass Spectrom. Rev. 2005, 24, 168–200.
  9. Wang, G.D. Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation & Applications; Cole, R.B., Ed.; John Wiley & Sons: New York, 1997, pp 137–74.
  10. Chernushevich, I.V.; Verentchikov, A.N.; Ens, W.; Standing, K.G. Effect of ion-molecule collisions in the vacuum chamber of an electrospray time-of-flight mass spectrometer on mass spectra of proteins. J. Amer. Soc. for Mass Spectrom. 1996, 7, 342–9.
  11. Amad, M.H.; Cech, N.B.; Jackson, G.S.; Enke, C.G. Importance of gas-phase proton affinities in determining the electrospray ionization response for analytes and solvents. J. Mass Spectrom. 2000, 35, 784–9.
  12. Iavarone, A.T.; Jurchen, J.C.; Williams, E.R. Effects of solvent on the maximum charge state and charge state distribution of protein ions produced by electrospray ionization. J. Amer. Soc. for Mass Spectrom. 2000, 11, 976–85.
  13. Sterner, J.L.; Johnston, M.V.; Nicol, G.R.; Ridge, D.P. Signal suppression in electrospray ionization Fourier transform mass spectrometry of multi-component samples. J. Mass Spectrom. 2000, 35, 385–91.
  14. Smith, R.D.; Loo, J.A.; Barinaga, C.J.; Edmonds, C.G.; Udseth, H.R. Collisional activation and collision-activated dissociation of large multiply charged polypeptides and proteins produced by electrospray ionization. J. Amer. Soc. for Mass Spectrom. 1990, 1, 53–65.
  15. Belov, M.E.; Gorshkov, M.V.; Alving, K.; Smith, R.D. Optimal pressure conditions for unbiased external ion accumulation in a two-dimensional radio-frequency quadrupole for Fourier transform ion cyclotron resonance mass spectrometry. Rap. Commun Mass Spectrom. 2001, 15, 1988–96.
  16. Schmidt, A.; Bahr, U.; Karas, M. Influence of pressure in the first pumping stage on analyte desolvation and fragmentation in nano-ESI MS. Anal. Chem. 2001, 73, 6040–6.
  17. Oberacher, H.; Walcher, W.; Huber, C.G. Effect of instrument tuning on the detectability of biopolymers in electrospray ionization mass spectrometry. J. Mass Spectrom. 2003, 38, 108–16.
  18. Vaidyanathan, S.; Broadhurst, D.I.; Kell, D.B.; Goodacre, R. Explanatory optimization of protein mass spectrometry via genetic search. Anal. Chem. 2003, 75, 6679–86.
  19. Vaidyanathan, S.; Kell, D.B.; Goodacre, R. Selective detection of proteins in mixtures using electrospray ionization mass spectrometry: influence of instrumental settings and implications for proteomics. Anal. Chem. 2004, 76, 5024–32.
  20. Michalewicz, Z.; Fogel, D.B. How to Solve it: Modern Heuristics. Springer-Verlag: Heidelberg, 2000.

Dr. Vaidyanathan is a Research Associate, School of Chemistry, the University of Manchester, Sackville St., Manchester M60 1QD, U.K.; tel.: +44 0161 306 4414; e-mail: [email protected].

Comments