Anal. Chem. 2001, 73, 6040-6046
Influence of Pressure in the First Pumping Stage on Analyte Desolvation and Fragmentation in Nano-ESI MS Andrea Schmidt,* Ute Bahr, and Michael Karas
Institute for Pharmaceutical Chemistry, J. W. Goethe University of Frankfurt, Marie-Curie-Strasse 9-11, 60439 Frankfurt, Germany
In ESI MS, some classes of biomolecules are detected only with low signal intensities due to difficulties in achieving efficient analyte desolvation, either because an analyte tends to fragment already at gentle desolvation conditions (i.e., noncovalent protein complexes or nucleotides) or because an analyte requires very strong activation in order to remove solvent molecules (i.e., carbohydrates). Even though the pressure in the first pumping stage of the ESI instrument is known to have an influence on the desolvation conditions, it has never been the focus of a detailed investigation. The role of the pressure in the first pumping stage is systematically interrogated in this study for several model substances. Ion signal intensities and signal-to-noise ratios are significiantly enhanced if the pressure in the first pumping stage is increased and adjusted, and analyte fragmentation can be substantially reduced. Thus, besides thermal heating and the acceleration in the nozzle-skimmer region, which are usually optimized, the pressure in the first pumping stage is an additional important desolvation parameter. In the past 10 years, electrospray ionization mass spectrometry (ESI MS)1-3 has become a powerful tool in bioanalytics and drug research and is today used as routine analytical method.4-6 In ESI MS, effective analyte desolvation is a prerequisite for the successful mass analysis of biomolecules. Solvent molecules that are bound unspecifically to the analyte have to be removed completely by energy input upon transfer of the ions from atmospheric pressure into the vacuum, by thermal or collisional activation, or both, while analyte dissociation has to be avoided. Fulfillment of these requirements becomes critical for analytes that tend to dissociate or fragment easily, such as noncovalently bound specific biomolecule assemblies (for reviews, see refs 7-9) or nucleotides.10 Furthermore, desolvation is difficult to achieve for analytes that are released with a low probability from the ESI* Corresponding author: (e-mail)
[email protected]; (fax) ++49 69 798 29918; (phone) ++49 69 798 29914. (1) Dole, M.; Mack, L. L.; Hines, R. L.; Mobley, R. C.; Ferguson, L. D.; Alice, M. B. J. Chem. Phys. 1968, 49, 2240-2249. (2) Aleksandrov, M. L.; Gall, L. N.; Krasnov, N. V.; Nikolaev, V. I.; Pavlenko, V. A.; Shkurnov, V. A. Dolk. Akad. Nauk SSSR 1984, 277, 379-383. (3) Yamashita, M.; Fenn, J. B. J. Phys. Chem. 1984, 88, 4451-4458. (4) Bakhtiar, R.; Nelson, R. W. Biochem. Pharmacol. 2000, 59, 891-905. (5) Pramanik, B. N.; Bartner, P. L.; Chen, G. Curr. Opin. Drug Discovery Dev. 1999, 2, 401-417. (6) Siuzdak, G.; Lewis, J. K. Biotechnol. Bioeng. 1998, 61, 127-134.
6040 Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
generated charged droplets due to their hydrophilicity and their lack of surface activity,11,12 such as carbohydrates.13 The analysis of noncovalent protein complexes or proteinligand adducts by ESI MS is still an analytical challenge, even though many examples of successful investigations have been published.7-9 One reason for this is that these complexes usually require the analysis from aqueous buffered solutions in order to avoid dissociation already in solution prior to mass analysis. This deviation from standard ESI conditions creates two problems: first, to generate a stable spray from this solvent and, second, to accomplish an effective ion desolvation. While the spray problems can be reduced by shaping metal spray capillaries14 or by using nano-ESI,13,15 a more demanding desolvation usually requires harsher interface conditions under which the labile complexes may be destroyed. Thus, a compromise in thermal or collisional energy input has to be made, and as a consequence, desolvation is usually incomplete, the ion signal intensities are low, and the peaks corresponding to the intact complex are often broadened due to adduct formation with solvent molecules or salt and buffer ions present in the analyte solution. Lower ion intensities compared to peptides and proteins are also observed for hydrophilic compounds, such as glycoproteins and oligosaccharides, which points to the important effect of the sample surface acitivity in ESI.11,12 Analytes present or even enriched on the droplet surface are preferred within the offspring droplet cascade16 while those in the interior of the droplet will be essentially lost. In acidic solutions, analytes such as peptides and proteins are highly charged. This charge may be even more relevant for their enrichment on the droplet surface and their easy and highly sensitive detection in all ESI configurations. The extent of desolvation by collisional activation is typically controlled by the acceleration potential in the first pumping stage (“nozzle-skimmer region”) with some additional fine-tuning in (7) Loo, J. A. Mass Spectrom. Rev. 1997, 16, 1-23. (8) Smith, R. D.; Bruce, J. E.; Wu, Q.; Lei, Q. P. Chem. Soc. Rev. 1997, 26, 191-202. (9) Veenstra, T. D. Biophys. Chem. 1999, 79, 63-79. (10) Nordhoff, E.; Kirpekar, F.; Roepstorff, P. Mass Spectrom. Rev. 1996, 15, 67-138. (11) Enke, C. G. Anal. Chem. 1997, 69, 4885-4893. (12) Cech, N. B.; Enke, C. G. Anal. Chem. 2000, 72, 2717-2723. (13) Bahr, U.; Pfenninger, A.; Karas, M.; Stahl, B. Anal. Chem. 1997, 69, 45304535. (14) Chowdhury, S. K.; Chait, B. T. Anal. Chem. 1991, 63, 1660-1664. (15) Wilm, M. S.; Mann, M. Int. J. Mass Spectrom. Ion Processes 1994, 136, 167180. (16) Kebarle, P. J. Mass Spectrom. 2000, 35, 804-817. 10.1021/ac010451h CCC: $20.00
© 2001 American Chemical Society Published on Web 11/15/2001
the skimmer-transfer quadrupole line. The influence of this parameter on the ESI mass spectra has been studied systematically,17 and tuning of the potential difference and the temperature of heated interface elements are general strategies in the adjustment of ESI MS desolvation conditions.18-23 Also, the pressure in the first pumping stage has a strong effect on the collisional activation and desolvation; nevertheless, mostly only rough numbers, i.e., approximately “1 mbar”, are given for the pressure in the nozzle-skimmer region.24 The increase of pressure in the first pumping stage has been reported to improve the quality of ESI mass spectra in the analysis of large protein complexes.25-28 A higher pressure was reported to reduce fragmentation and to yield higher signal intensities of these fragile compounds. Even though the influence of pressure in the first pumping stage on the transmission efficiency of ions29 or on the desolvation conditions19,30 was mentioned previously, this issue was not studied systematically until today. Here we report about investigations on the influence of the nozzle-skimmer pressure by using noncovalent complexes, carbohydrates, and nucleotides as model substances. It will be shown that adjusting the pressure is an important experimental parameter needed in order to optimize ion generation in ESI. EXPERIMENTAL SECTION ESI MS experiments were performed on an orthogonalacceleration time-of-flight instrument (oTOF MS Mariner, PE/ PerSeptive Biosystems, Framingham, MA) equipped with a nanoESI source (Protana, Odense, Denmark). The atmospheric pressure-vacuum interface was modified from the original nozzle-curtain gas setup to a heated metal-transfer capillary setup (150 mm in length, 0.5-mm i.d.) surrounded by a spray chamber. Furthermore, a needle valve was installed in the source flange in order to allow pressure adjustment in the first pumping stage by (17) Smith, R. D.; Loo, J. A.; Barinaga, C. J.; Edmonds, C. G.; Udseth, H. R. J. Am. Soc. Mass Spectrom. 1990, 1, 53-65. (18) Chowdhury, S. K.; Katta, V.; Chait, B. T. Rapid Commun. Mass Spectrom. 1990, 4, 81-87. (19) Smith, R. D.; Light-Wahl, K. J. Biol. Mass Spectrom. 1993, 22, 493-501. (20) Light-Wahl, K. J.; Schwartz, B. L.; Smith, R. D. J. Am. Chem. Soc. 1994, 116, 5271-5278. (21) Schwartz, B. L.; Bruce, J. E.; Anderson, G. A.; Hofstadler, S. A.; Rockwood, A. L.; Smith, R. D.; Chilkoti, A.; Stayton, P. S. J. Am. Soc. Mass Spectrom. 1995, 6, 459-465. (22) Loo, J. A. J. Mass Spectrom. 1995, 30, 180-183. (23) Smith, R. D.; Schwartz, B. L.; Gale, D. C. In Methods in Molecular Biology: Protein and Peptide Analysis by Mass Spectrometry, Chapman, J. R., Ed.; Humana Press: Totowa, NJ, 1996; pp 115-127. (24) Bruins, A. P. In Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation and Applications; Cole, R., Ed.; John Wiley & Sons: New York, 1997; pp 107-136. (25) Potier, N.; Barth, P.; Tritsch, D.; Biellmann, J.-F.; Van Dorsselaer, A. Eur. J. Biochem. 1997, 243, 274-282. (26) Strupat, K.; Carte, N.; Rogniaux, H.; Leize, E.; Van Dorsselaer, A.; Proceedings of the 47th ASMS Conference on Mass Spectrometry and Allied Topics, Dallas, TX, 1999; pp 1287-1288. (27) Rogniaux, H.; Van Dorsselaer, A.; Barth, P.; Biellmann, J.-F.; Barbanton, J.; van Zandt, M.; Chevrier, B.; Howard, E.; Mitschler, A.; Potier, N.; Urzhumtseva, L.; Moras, D.; Podjarny, A. J. Am. Soc. Mass Spectrom. 1999, 10, 635647. (28) Van Berkel, W. J. H.; Van den Heuven, R. H. H.; Versluis, C.; Heck, A. J. R. Protein Sci. 2000, 9, 435-439. (29) Smith, R. D.; Barinaga, C. J. Rapid Commun. Mass Spectrom. 1990, 4, 5457. (30) Wang, G.; Cole, R. B. In Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation and Applications; Cole, R., Ed.; John Wiley & Sons: New York, 1997; pp 137-174.
Figure 1. Schematic view of the atmospheric pressure-vacuum interface (in-house modified PE/PerSeptive Biosystems Mariner oTOF MS). Parameter ranges used: Tcap, 80-200 °C; Ucap, 12-300 V; Usk1, 6 V. Pressure p1 was measured by an additional gauge and was varied between 0.86 (needle valve closed) and 4 mbar.
either the inlet of nitrogen, argon, or helium, Figure 1. The pressure can be measured by built-in gauges in the second pumping stage, i.e., in the transfer quadrupole region (p2, Pirani gauge) and in the mass analyzer (ptof, cold cathode gauge). An additional Pirani gauge was installed to determine the pressure in the first pumping stage (p1). The pressure p1 was varied between 0.84 (closed valve) and 4 mbar, corresponding to p2 between 6.5 × 10-3 and 27.9 × 10-3 mbar. Desolvation of the ions generated via nano-ESI was accomplished by thermal activation in the heated transfer capillary (Tcap, 80-200 °C) and by collisions with the residual gas in the first pumping stage, caused by a potential difference applied between the transfer capillary (Ucap, 10-300 V) and the first skimmer (Uski, 6-12 V). Nano-ESI needles were pulled from borosilicate glass capillaries (Clark Electromedical Instruments, Pangbourne, U.K., 1.2mm o.d., 0.65-mm i.d.) with a micropipet puller (model P-97, Sutter Instrument Co., Novato, CA) to a fine tip with a final inner diameter of approximately 1-10 µm and were coated with gold by means of a sputter coater (model K 550, Emitech, Ashford, U.K.). Maltopentaose (Sigma, Deisenhofen, Germany) was solved in water/methanol (1:1), horse heart myoglobin (Sigma, Deisenhofen, Germany) was solved in water, and alcohol dehydrogenase (Serva Feinbiochemica, Heidelberg, Germany), concanavalin A (Sigma, Deisenhofen, Germany), and a DNA 24-mer (with mixed sequence, composition d(A8 C6 G4 T6), Sigma ARK, Darmstadt, Germany) were solved in 20 mM ammonium acetate (Fluka Chemie AG, Buchs, Switzerland) adjusted with ammonia to a final pH of 8.5. All analyte solutions were prepared in a final concentration between 5 and 10 pmol/µL. A 3-µL aliquot of each solution were loaded into the nano-ESI spray needles. RESULTS AND DISCUSSION General Considerations: Maltopentaose as a Model Compound. In comparison to peptides and proteins, underivatized carbohydrates generally require harsher desolvation conditions in order to obtain mass spectra with acceptable ion signal intensity, i.e., a higher transfer capillary temperature and a higher capillary voltage (higher potential between transfer capillary and skimmer) have to be applied. Hydrophilic compounds, such as carbohydrates, exhibit only weak affinity toward the surface of the charged Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
6041
Figure 2. Nano ESI MS of maltopentaose, sum of 10 spectra (3 s each) at Tcap ) 200 °C and Ucap ) 200 V: (a) needle valve closed, p1 ) 0.84 mbar; (b) inlet of nitrogen, p1 ) 2.0 mbar.
initial electrospray droplets. Thus, their transfer through the droplet fission cascade and their release into the gas phase generally require a higher extent of solvent evaporation and thus a higher extent of energy input. In this context, it is interesting to note that even under gentle desolvation conditions when desolvation is expected to be incomplete distinct adducts between solvent molecules and the analyte are seldomly reported. Analyte fragmentation is a process competitive to analyte desolvation under the conditions required for the analysis of hydrophilic componds, as shown in Figure 2a for maltopentaose.
Besides the singly charged sodiated molecular ion [M + Na]+ at m/z 851.1, a series of fragment ions are detected upon collisional activation in the first pumping stage, the most abundant at m/z 689.0 corresponding to the loss of a glucose unit from the intact molecular ion ([M - 162 Da + Na]+). If the pressure in the desolvation zone is raised from 0.84 to 2.0 mbar by admission of N2 gas through the needle valve while all other desolvation parameters are kept constant, the mass spectra of maltopentaose undergo substantial changes (Figure 2b). Analyte fragmentation is significiantly reduced, and a higher absolute ion signal intensity and a higher signal-to-noise ratio are obtained. The dependence of signal intensities of the molecular ion [M + Na]+ and the representative fragment ion [M - 162 Da + Na]+ on the variation of Ucap at a pressure p1 ) 0.84 mbar is shown in Figure 3a. The signal intensity of the molecular ion increases with increasing capillary potential and thus harsher desolvation conditions until analyte fragmentation becomes the predominant process. As a consequence, the ratio between the signal intensities of the fragment ion and the molecular ion increases steeply once a certain Ucap value is exceeded (Figure 3b). If the pressure p1 is increased at constant Ucap ) 200 V, fragmentation is reduced and the molecular ion dominates the mass spectra, as shown in Figure 3d. Maximum ion signal intensity is obtained at p1 ≈ 2 mbar; at a higher pressure, the signal intensity decreases again (Figure 3c). Variation of pressure p1 up to a distinct limit also induces an increase in absolute ion signal intensity; for example, the maximum signal intensity obtained at p1 ) 1.22 mbar is lower by a factor of ∼2 than the one obtained at p1 ) 2.37 mbar (with Ucap optimized, respectively; data not shown). The improvement in signal intensity may be
Figure 3. Maltopentaose at different desolvation conditions, Tcap ) 200 °C: (a) influence of Ucap on the intensity of the sodium ion attached molecule and the fragment [M - 162 + Na]+ (m/z 689) at p1 ) 0.84 mbar; (b) intensity ratio [fragment]/[molecular ion] at variation of Ucap, p1 ) 0.84 mbar; (c) influence of pressure p1 on the intensity of the [M + Na]+ ion and the fragment ion, Ucap ) 200 V; (d) intensity ratio [fragment]/ [molecular ion] at variation of p1, Ucap ) 200 V. 6042 Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
Figure 4. Influence of p1 (corrected for the different gases) on the intensity ratio [fragment ion]/[molecular ion] for maltopentaose and the prominent fragment ion at m/z 689, Ucap ) 200 V, and Tcap ) 200 °C. Dotted line, inlet of helium; solid line, inlet of nitrogen; dashed line, inlet of argon.
counterweighted by ion scattering upon the collisions with residual gas if the pressure in the first pumping stage is too high; due to this effect, the number of ions reaching the mass analyzer is reduced and the signal intensity decreases again (Figure 3c). A pressure rise in p1 also increases the pressure in the second pumping stage (p2) and in the mass analyzer (ptof) in the instrument used because it is only equipped with one turbomolecular pump. It is known that mass spectral resolution decreases with increasing pressure in the TOF analyzer due to ion scattering and ion fragmentation.31 In our experiments, we observed a reduction in resolution of ∼15% because ptof increases to 2.7 × 10-6 mbar if p1 is raised to 2.5 mbar, i.e., thereby setting a practical limit for the pressure range investigated. Regarding p2, it has been shown previously that an increase of pressure in the transfer quadrupole region has a strong beneficial effect on ion transmission and thus on ion signal intensity.32 Thus, in our experiments, two effects can be discussed as a rationale for the observed gain in ion signal intensity upon pressure increase. The higher ion signal intensity might be either due to improved analyte desolvation in the first pumping stage or due to enhanced ion transmission efficiency through the transfer quadrupole. As our instrument does not allow a variation of pressure exclusively in the first or second pumping stage, it is impossible to distinguish between the contributions of both effects on the resulting mass spectra. However, the changes observed in the extent of analyte fragmentation are indicative for changes in ion activation conditions (and thus desolvation conditions) upon pressure variaton. The results obtained in the investigation of maltopentaose as a model substance (Figure 3) can be rationalized by considering the changes in energy input due to collisions that are connected with pressure variation in the first pumping stage. If the pressure in the region between transfer capillary and first skimmer is raised, the mean free path λ decreases as λ is proportional to 1/pressure.33 Thus, the number of collisions an ion undergoes while passing (31) Chernushevich, I.; Verentchikov, A.; Standing, K. G.; Ens, W. J. Am. Soc. Mass Spectrom. 1996, 7, 342-349. (32) Douglas, D. J.; French, J. B. J. Am. Soc. Mass Spectrom. 1992, 3, 398-408. (33) Atkins, P. Physikalische Chemie, 1st ed.; VCH Verlagsgesellschaft: Weinheim, 1990.
Figure 5. Nano ESI MS of myoglobin, Tcap ) 150 °C, sum of 10 spectra (3 s each): (a) Ucap ) 150 V, p1 ) 1.36 mbar; signals labeled with solid circles, intact holomyoglobin; open circles, apomyoglobin. (b) Increase of p1 to 3.05 mbar. (c) Intensity ratio of the apomyoglobin signal at m/z 1696 and the holomyoglobin signal at m/z 1757 upon variation of p1. Dotted line, Ucap ) 150 V; dashed line, Ucap ) 200 V.
the first pumping stage increases in first approximation by the same factor as the pressure increase. The maximum energy transferred upon an inelastic collision is given by the center-of-mass energy (eq 1), with m1 and m2 the
E ) (1/2)vrel2(m1m2)/(m1 + m2)
(1)
masses of the ion and the residual gas molecule, respectively, and vrel ) vgas - vion the relative velocity of both species.34 The ions are accelerated in the electric field between transfer capillary and first skimmer, and the velocity they reach prior to a collision is proportional to the square root of the acceleration distance, v ∼ s1/2. Thus, due to the decrease in the mean free path, the acceleration distance s between two collisions decreases (as s ) λ) and the center-of-mass energy E is reduced by 1/(factor of pressure increase). Besides the influence of the center-of-mass energy for the collisions, also changes in the time period for analyte ion desolvation should be considered. The lower maximum ion velocity at higher pressure (for any residual gas) increases (34) Moonen, F.; Collette, C.; DePauw, E. In New Methods for the Study of Biomolecular Complexes; Ens, W., Standing, K. G., Chernushevich, I. V., Eds.; NATO ASI Series C: Mathematical and Physical Sciences 510; Kluwer Academic Publishers: Dordrecht, The Netherlands, 1998; pp 157-169.
Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
6043
Figure 6. Nano ESI MS of concanavalin A in ammonium acetate (pH 8.5), Tcap ) 80 °C, sum of 20 spectra (4 s each); left panel, needle valve closed; right panel, inlet of nitrogen: (a) Ucap ) 12 V, p1 ) 1.7 mbar; (b) Ucap ) 300 V, p1 ) 1.7 mbar; (c) Ucap ) 12 V, p1 ) 4.1 mbar; (d) Ucap ) 300 V, p1 ) 4.1 mbar.
the time the ions spend in the acceleration region by the factor given by the pressure increase. In summary, the ions undergo more but less energetic collisions with the residual gas if the pressure is raised at constant Ucap, and the time scale for collisional activation is extended. According to eq 1, the molecular mass of the residual gas also determines the center-of-mass energy of the collisions, and gases of higher molecular mass may enhance the fragmentation yield in CID experiments.35 However, if the composition of the residual gas in the first pumping stage is varied indirectly by using different curtain gases in the interface region, condensation and cooling effects upon the free jet expansion into the first pumping stage may alter the actual collisional energy input in the ESI-generated ions. As a consequence, a lower extent of fragmentation is observed upon the use of argon as a curtain gas compared to the use of nitrogen.36 To investigate the influence of the molecular weight of the residual gas, we compared the mass spectra obtained for maltopentaose at different p1 upon the inlet of nitrogen (Mr ) 28.01) with those obtained upon the inlet of helium (Mr ) 4.00) and argon (Mr ) 39.95) (Figure 4). As the pressure measured by means of (35) Douglas, D. J. J. Phys. Chem. 1982, 86, 185-191. (36) Schneider, B. B.; Douglas, D. J.; Chen, D. D. Y. Rapid Commun. Mass Spectrom. 2001, 15, 249-287.
6044 Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
a Pirani gauge is gas dependent, p1 values for argon and helium are corrected corresponding to the calibration curves given in the gauge’s manual.37 Significant differences can be detected in comparing the [fragment ion]/[molecular ion] ratios in dependence of the pressure p1 for the three different gases. If helium is used as the residual gas, fragmentation is reduced already at a lower pressure p1 compared to nitrogen, while a higher pressure p1 is required in the case of argon. These differences can be rationalized by considering the changes in the collisional energy input upon variation of the residual gas. At a given pressure and a given acceleration potential, the center-of-mass energy (eq 1) for argon and therefore the maximum energy input upon a collision is 1.4 times higher than that for nitrogen due to the differences in molecular mass of the residual gases. Thus, fragmentation of the molecular ion is enhanced in the case of argon. If the residual gas is helium instead of nitrogen, the centerof-mass energy is lower by a factor of 0.15 and therefore the extent of fragmentation is lower. The pressure-dependent ratios shown in Figure 4 reflect these differences only qualitatively, pointing to the fact that also other parameters (i.e., condensation effects36) play an important role in determining the actual desolvation conditions in the first pumping stage. Even though in our (37) Balzers Compact Pirani Gauge TPR 250 Manual.
investigations helium exhibits the strongest potentials in reducing analyte fragmentation, a similar extent of reduction and therefore improvement in absolute ion signal intensity and signal-to-noise ratio can be obtained with nitrogen at a higher but still practically useful pressure p1. Noncovalent Protein Complexes. To study the beneficial effect of pressure rise on the analysis of noncovalent complexes systematically, we chose myoglobin as a model system. As in myoglobin the bond between the protein and the prosthetic heme group is noncovalent and therefore weak, the complex dissociates easily upon collisional activation and alterations in the activation conditions can be monitored by the relative ion signal intensities of the complex and the apoprotein chain. At a pressure p1 of 1.36 mbar, the intact heme-bound complex (holomyoglobin) is observed with dominant signal intensity only at a low capillary potential (Ucap ) 50 V). If Ucap is raised, fragmentation occurs and the signals of the heme group at 616.2 Th and of apomyoglobin are predominant, as shown in Figure 5a for Ucap ) 150 V. An increase of p1 to 3.05 mbar by the admission of nitrogen under these conditions reduces fragmentation substantially (Figure 5b). The influence of variation of p1 on the fragmentation of holomyoglobin is shown in Figure 5c for the ratio between two representative ion signals (charge state +10; m/z 1696 Th for apomyoglobin and m/z 1757 Th for holomyoglobin). The signal intensity of holomyoglobin increases relative to the apomyoglobin signals with increasing pressure p1, and fragmentation is almost absent at p1 above 2.2 mbar. It is notable that this effect is more pronounced for a higher capillary potential, i.e., Ucap ) 200 V. The effect of pressure p1 on the detection of large protein complexes can be demonstrated with concanavalin A, a protein that forms specific multimers depending on the pH of the analyte solution. At a pH below 5 only monomers are present in solution, at a pH around 5.5 specific dimers are formed, and at a pH above 7 a tetramer with a molecular weight of ∼102 000 exists in solution.38 As has been shown previously, the intact complex can be observed in ESI MS, and also the pH-dependent association has been studied by mass spectrometry.38-40 Under gentle desolvation conditions, i.e., a very low capillary temperature, a very low capillary potential, and a pressure p1 ) 1.7 mbar, the ion signals of the intact tetramer dominate the ESI mass spectra in positive ion mode (Figure 6a). However, the signals are broad and are observed in low intensity, as reflected by the poor signalto-noise ratio. Increasing the capillary potential helps to decluster the analyte ions and produces more narrow peaks with higher signal-to-noise ratio, but also leads to dissociation of the tetramer into the monomer subunits (Figure 6b). If the pressure p1 is raised to 4.1 mbar, the quality of the mass spectra obtained under gentle desolvation conditions improves with respect to signal width, signal intensity, and signal-to-noise ratio (Figure 6c). As discussed above, this can be due to changes in desolvation conditions as well as changes in ion transmission efficiency through the transfer quadrupole as p2 is also changed. (38) Light-Wahl, K. J.; Winger, B. E.; Smith, R. D. J. Am. Chem. Soc. 1993, 115, 5869-5870. (39) Light-Wahl, K. J.; Schwartz, B. L.; Smith, R. D. J. Am. Chem. Soc. 1994, 116, 5271-5278. (40) Wang, J.; Busman, M.; Knapp, D. R. Proceedings of the 44th ASMS Conference on Mass Spectrometry and Allied Topics, Portland, OR 1996; p 1063.
Figure 7. Nano ESI MS of a DNA 24mer with mixed sequence, Tcap ) 180 °C, Ucap ) 50 V, sum of 25 spectra (5 s each): (a) needle valve closed, p1 ) 1.14 mbar; (b) inlet of nitrogen, p1 ) 2.99 mbar; (c) intensity ratio of the [M - 12H]12-/[M - 5H]5- charge states at variation of p1.
Besides the changes in ion signal intensity, the analyte ions tolerate harsher desolvation conditions without fragmentation into subunits, as described previously.27 This is shown as an example for a capillary potential of Ucap ) 300 V in Figure 6d and can be also observed for an increase in capillary temperature or upon collisional activation between skimmer and transfer quadrupole (data not shown). Thus, the increase of pressure allows tuning of desolvation conditions over a wider range while preserving the intact noncovalent complex. The optimization of pressure as an additional desolvation parameter helps to improve ion signal intensity and signal-to-noise ratio for noncovalent complexes, especially those with high molecular weight. Alcohol dehydrogenase, a tetrameric complex with Mr ≈ 148 000 consisting of four identical subunits can be detected with ∼10-fold signal intensity if p1 is raised from 0.84 to 2.45 mbar (data not shown). Furthermore, fragmentation into the monomer subunits is substancially reduced. Nucleotides. Nucleotides require a careful adjustment of the ESI desolvation conditions in order to compromise between the efficient removal of unspecific adducts and the prevention of fragmentation of the molecular ion. Even though analyte lability is not the major problem in the analysis of nucleotides by ESI MS but rather the pronounced metal ion attachment, it is known that the tendency of nucleotides toward fragmentation increases Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
6045
with increasing number of bases.10 Thus, we studied whether pressure variation also had a beneficial effect on the analysis of these compounds. A synthetic DNA 24-mer with mixed sequence (Mr ) 7318) was chosen as a model analyte; ESI MS experiments were performed in negative ion mode. At a pressure p1 of 1.14 mbar and gentle desolvation conditions, the DNA 24-mer can be detected in the charge states - 4 to - 12 (Figure 7a). The application of only slighty harsher conditions immediately leads to strong fragmentation and thus to a decrease in molecular ion signal intensity. If the pressure p1 is raised, the overall ion signal intensity of all charge states increases almost linearly (data not shown), and signal-to-noise ratio increases (Figure 7b for p1 ) 2.99 mbar). The higher pressure p1 also permits the variation of the other desolvation parameters (Ucap, Tcap) over a wider range without inducing fragmentation of the molecular ions, as shown for carbohydrates and noncovalent complexes above. Thus, the optimization of the ESI MS conditions for the analysis of nucleotides becomes less critical if the pressure p1 is raised. It is interesting to note that the pressure increase also causes a shift in the bimodal charge-state distribution toward the lower charged molecular ions (Figure 7c). A shift in the charge-state distribution of DNA has been observed upon addition of imidazole41 or different concentrations of ammonium acetate42 to the analyte solution and was rationalized by structural alterations of the DNA oligomer in solution. However, in the present study, a similar shift is observed although identical analyte solutions were used for the series of experiments. Thus, in this case, chargestate reduction cannot be caused by solution-phase processes but may be due to proton-transfer reactions occurring upon transfer of the nucleotide in the gas phase. According to the established models of the ESI mechanism,43 gas-phase proton-transfer occurs in a cluster between the analyte (i.e., DNA) and counterions (i.e.,
ammonium ions). Since ions of lower charge states are observed if the pressure in the first pumping stage is raised, it can be concluded that at a higher p1 more counterions are present in the final cluster, transferring more protons to the highly negative charged analyte ion. The changes in charge-state distribution at higher p1 may therefore point to a more effective access to ions from residual larger droplets in the offspring cascade which are usually suppressed or lost.
(41) Greig, M.; Griffey, R. H. Rapid Commun. Mass Spectrom. 1995, 9, 97-102. (42) Griffey, R. H.; Sasmor, H.; Greig, M. J. J. Am. Soc. Mass Spectrom. 1997, 8, 155-160.
(43) Kebarle, P.; Ho, Y. In Electrospray Ionization Mass Spectrometry: Fundamentals, Instrumentation and Applications; Cole, R., Ed.; John Wiley & Sons: New York, 1997; pp 3-64.
6046
Analytical Chemistry, Vol. 73, No. 24, December 15, 2001
CONCLUSION The pressure in the first pumping stage has a significant influence on the quality of mass spectra obtained in nano-ESI MS. Especially in the case of compounds that require comparably harsh desolvation conditions or in the case of labile compounds that tend to fragment upon desolvation, an increase of pressure helps to desolvate the analyte ions. At the same time, a rise in pressure reduces fragmentation substantially. Thus, a higher analyte ion signal intensity can be obtained, as shown for carbohydrates, DNA, and noncovalent complexes in this study, and the signal-to-noise ratio in the mass spectra is significiantly improved. The pressure in the first pumping stage has to be considered as an additional important desolvation parameter. Adjustment of other desolvation parameters such as the acceleration potential between nozzle (heated transfer capillary) and skimmer or the capillary temperature is less critical if the pressure is raised. This is especially important for “delicate” compounds and is a prerequisite for their routine analysis by ESI MS.
Received for review April 19, 2001. Accepted October 3, 2001. AC010451H
