Methods for Generating Protein Molecular Ions in ToF-SIMS

Mar 23, 2004 - Sally L. McArthur,* Marie C. Vendettuoli, Buddy D. Ratner, and. David G. Castner. National ESCA and Surface Analysis Center for Biomedi...
1 downloads 0 Views 111KB Size
3704

Langmuir 2004, 20, 3704-3709

Methods for Generating Protein Molecular Ions in ToF-SIMS Sally L. McArthur,* Marie C. Vendettuoli, Buddy D. Ratner, and David G. Castner National ESCA and Surface Analysis Center for Biomedical Problems (NESAC/Bio), Departments of Bioengineering and Chemical Engineering, University of Washington, Box 351750, Seattle, Washington 98195-1750 Received October 2, 2003. In Final Form: February 7, 2004 One of the greatest challenges in mass spectrometry lies in the generation and detection of molecular ions that can be used to directly identify the protein from the molecular weight of the molecular ion. Typically, proteins are large (MW > 1000), nonvolatile, and/or thermally labile, but the vaporization process produced by many mass spectrometry techniques including time-of-flight secondary ion mass spectrometry (ToF-SIMS) is inherently limited to generating ions from smaller compounds or fragments of the parent molecule, making the identification of proteins complex. The application of specific molecules to aid in the generation of high molecular weight ions in ToF-SIMS has been recognized for some time. In this study we have developed a matrix-SAM substrate preparation technique based on the self-assembly of a matrix-like molecule, mercaptonicotinic acid (MNA), on gold. We then compare this substrate with two existing ToF-SIMS sample preparation techniques, cationized alkane thiol and matrix-enhanced SIMS (MESIMS). The results of this study illustrate that while there is a range of methods that can be used to improve the molecular ion yield of proteins in ToF-SIMS, their efficacy and reproducibility vary considerably and crucially are linked to the sample preparation and/or protein application methods used. Critically, the MNA modified substrate was able to simultaneously induce molecular ions for each protein present in a multicomponent solution, suggesting that this sample preparation technique may have future application in proteomics and DNA analysis.

Introduction One of the greatest challenges in mass spectrometry lies in the development of organic sample preparation routes that lead to an improvement of the instrument performance. This is particularly critical in time-of-flight secondary ion mass spectrometry (ToF-SIMS), where methods that enhance the useful signal, access a broader mass range, and reduce molecule fragmentation are required.1 Molecule fragmentation is particularly problematic when analyzing biological samples using mass spectrometry. Typically, biomolecules such as proteins are large (MW > 1000), nonvolatile, and thermally labile. The vaporization process used in ToF-SIMS is predisposed toward generating ions from smaller compounds or fragments of the parent molecule. An example of this is seen in the analysis of adsorbed protein films where the largest ions detected are the immonium ions (+NH2dCHR) from each amino acid (MW < 200).2As a result of fragmentation, the identification of proteins is often more like a jigsaw puzzle, where the amino acid fragments have to be pieced together using pattern recognition or multivariate analysis techniques, to identify the parent molecules.3,4 Recent developments in SIMS ion sources have shown that polyatomic or cluster ion sources can significantly * Corresponding author. Current address: Department of Engineering Materials, University of Sheffield, Mappin St Sheffield, United Kingdom S1 3JD. Phone, (44) 114 2225513; fax, (44) 114 2225943; e-mail, [email protected]. (1) Delcorte, A.; Medard, N.; Bertrand, P. Anal. Chem. 2002, 74, 4955. (2) Mantus, D. S.; Ratner, B. D.; Carlson, B. A.; Moulder, J. F. Anal. Chem. 1993, 65, 1431. (3) Wagner, M. S.; Castner, D. G. Langmuir 2001, 17, 4649. (4) Wagner, M. S.; Tyler, B. J.; Castner, D. G. Anal. Chem. 2002, 74, 1824.

improve the molecular ion yield of both biological and polymeric materials with MW < 2000.5-7 While the availability of both C60 and polyatomic gold sources has increased dramatically in the last year, existing instrumentation and sources still require the development of alternate routes for molecular ion detection. Historically, the detection of larger organic fragments from polymers and proteins in ToF-SIMS has been achieved through metal ion cationization. By placing the analyte on an etched silver substrate, researchers have been able to expand the effective mass range of SIMS for submonolayer organic molecules to 10 000 m/z.8 Surface metallization with 2-60 nmol of gold/cm2 has also been shown to be an effective route for improving molecular ion yield in polymer samples with MW < 3000.1 While the technique is fundamentally an inversion of the silver foil cationization route, it has the added advantage of being applicable to any type of sample, not just those that can be dissolved and cast as thin films. Other variations on the cationization theme have included the incorporation of saturated solutions of ammonium or sodium chloride with the analyte prior to deposition onto metallic substrates9,10 and the deposition of a submonolayer of cocaine hydrochloride onto a gold substrate prior to deposition of the analyte.11 Recent studies have also shown that metal ion substituted carboxyl terminated self-assembled monolayers act as both effective cationization substrates and as model systems (5) Davies, N.; Weibel, D. E.; Blenkinsopp, P.; Lockyer, N.; Hill, R.; Vickerman, J. C. Appl. Surf. Sci. 2003, 203-204, 223. (6) Wong, S. C. C.; Hill, R.; Blenkinsopp, P.; Lockyer, N. P.; Weibel, D. E.; Vickerman, J. C. Appl. Surf. Sci. 2003, 203-204, 219. (7) Gillen, G.; Fahey, A. Appl. Surf. Sci. 2003, 203-204, 209. (8) Benninghoven, A. Surf. Sci. 1994, 300, 246. (9) Liu, L. K.; Busch, K. L.; Cooks, R. G. Anal. Chem. 1981, 53, 109. (10) Parees, D. M.; Hanton, S. D.; Clark, P. A. C.; Willcox, D. A. J. Am. Soc. Mass Spectrom. 1998, 9, 282. (11) Nicola, A. J.; Muddiman, D. C.; Hercules, D. M. J. Am. Soc. Mass Spectrom. 1996, 7, 467.

10.1021/la0358419 CCC: $27.50 © 2004 American Chemical Society Published on Web 03/23/2004

Protein Molecular Ions in ToF-SIMS

for investigating the efficacy of the cationization process with different proteins and polymers.12,13 The application of specific molecules, such as cationization salts, for improving molecular ion intensity in mass spectrometry has been recognized for some time. Matrixassisted laser desorption ionization mass spectrometry (MALDI-MS) uses small, UV-absorbing molecules (typically an organic acid) to modulate the ionization process. The absorption of laser irradiation by the matrix leads to rapid evaporative ejection of material. Analyte molecules are embedded in the photoabsorbing matrix and photoactivated reactions lead to ionization of both the matrix and analyte molecules with minimal or no fragmentation.14 Fundamentally, the saturated ammonium or sodium chloride solutions discussed earlier act as a matrix in the SIMS process. In the mid 1990s, attempts to better understand the ionization process and the effects of sample preparation in MALDI-MS led to the investigation of MALDI-MS matrixes in ToF-SIMS sample preparation.15,16 By combining a common matrix dihydroxybenzoic acid (DHB) with the analyte, Wu and Odom developed matrix-enhanced SIMS (MESIMS), a technique that enabled them to generate molecular ions from a range of proteins and polypeptides with MW < 10 000 m/z.15 One of the limitations in using a MALDI-MS matrix for ToF-SIMS studies lies in the nature of the crystals formed during sample drying and the surface sensitivity of the static SIMS technique. A number of studies have investigated the structure and location of analyte molecules within MALDI-MS matrix crystals.16-19 Using fluorescence and transmission microscopy, Dai et al. were able to image the location of fluorescently labeled proteins within a range of matrix crystals and demonstrated that while the protein does tend to associate with the matrix crystal, it can be embedded deep within the structure.17 Other factors including solvent evaporation rates, analyte solubility, substrate characteristics and the location of the analyte into the matrix crystals have been shown to affect the success of MALDI-MS analysis.16,18,19 While ions detected in the MALDI-MS process likely come from a relatively large volume, the molecular secondary ions sputtered under the keV conditions present in static SIMS can by definition only originate from the first few nanometers of the surface.15,20 Sample preparation and thus the size of the matrix crystals and location of the analyte within them becomes a critical factor in the success of MESIMS. The actual mechanisms involved in the desorption and ionization processes and the exact role of the matrix in MALDI-MS are still relatively poorly understood.18,21One method that has been developed to eliminate some of the variability related to the matrix crystallization has been the self-assembly of the matrixlike molecules as monolayers (SAMs) on a gold substrate prior to addition of the analyte.21 This process has advantages as a sample (12) Michel, R.; Luginbuhl, R.; Graham, D. J.; Ratner, B. D. J. Vac. Sci. Technol. A. 2000, 18, 1114. (13) Michel, R.; Luginbuhl, R.; Graham, D. J.; Ratner, B. D. Langmuir 2000, 16, 6503. (14) Ehring, H.; Sundqvist, B. U. R. J. Mass Spectrom. 1995, 30, 1303. (15) Wu, K. J.; Odom, R. W. Anal. Chem. 1996, 68, 873. (16) Hanton, S. D.; Clark, P. A. C.; Owens, K. G. J. Am. Soc. Mass Spectrom. 1999, 10, 104. (17) Dai, Y.; Whittal, R. M.; Li, L. Anal. Chem. 1996, 68, 2495-2500. (18) Horneffer, V.; Driesewerd, K.; Ludmann, H.-C.; Hillenkamp, F.; Lage, M.; Strupat, K. Int. J. Mass Spectrom. 1999, 185-187, 859. (19) Garden, R. W.; Sweedler, J. V. Anal. Chem. 2000, 72, 30. (20) Wittmaack, K.; Szymczak, W.; Hoheisel, G.; Tuszynski, W. J. Am. Soc. Mass Spectrom. 2000, 11, 553. (21) Mouradian, S.; Nelson, C. M.; Smith, L. M. J. Am. Chem. Soc. 1996, 118, 8639.

Langmuir, Vol. 20, No. 9, 2004 3705

preparation technique for MALDI-MS as it ensures that the analyte remains at the surface of the sample and within the analysis range. While there have been no published reports of using a similar approach to induce matrixenhanced ToF-SIMS of proteins, previous studies have successfully used aminoethanethiol SAMs as a substrate for the ToF-SIMS analysis of small (MW < 300) organosulfates and sulfonates molecules.22 In this study we have developed a matrix-SAM substrate preparation technique based on the self-assembly of a matrixlike molecule, mercaptonicotinic acid (MNA), on gold. We then compare this substrate with two existing ToF-SIMS sample preparation techniques, cationized alkane thiol and MESIMS. The results of this study illustrate that while there is a range of methods that can be used to improve the molecular ion yield of proteins in ToF-SIMS, their efficacy and reproducibility vary considerably and are crucially linked to the sample preparation and/or protein application methods used. Most importantly, the MNA modified substrate was able to simultaneously induce molecular ions for each protein present in a multicomponent solution, suggesting that this sample preparation technique may have future application in proteomic and DNA analysis. Materials and Methods Materials. Prior to treatment, silicon wafer fragments were sonicated in methylene chloride followed by ethanol. They were then dried rapidly under a stream of nitrogen gas and stored in until use in a nitrogen atmosphere. Gold coatings were prepared by first evaporating an adhesion layer of chromium onto clean silicon wafers followed by the layer of gold (∼3000 Å of Au/∼ 200 Å of Cr). Sigma-Aldrich (Milwaulkee, WI) supplied all chemicals used except where stated. All chemicals were of the highest purity available. Sample Preparation. Illustrations of the three sample preparation routes used in the study are shown in Figure 1. Matrix-Enhanced SIMS (MESIMS). Samples were prepared following the standard dried droplet methods used in MALDI-MS.23 The matrix (dihydroxybenzoic acid, DHB or sinapinic acid, SA) was dissolved as a saturated solution in a 50/50 vol % of acetonitrile (ACN) and Milli Q grade water with the addition of 0.1 vol % of trifluoroacetic acid. In some instances the solution was then passed through Alltech (Extract Clean 200 mg/3 mL) cation-exchange resin columns to remove sodium ions. Samples were prepared by mixing the protein (0.1 mg/mL in water) and matrix (as supplied or desalted) in a 1:1 ratio. Droplets of the protein/matrix solution, 0.5-1.5 µL, were then applied to clean silicon and gold coated wafers and dried at room temperature. To assess the effect of sample application, samples were also prepared by spin coating the matrix/protein solution onto the wafers (details below). Cationized Self-Assembled Monolayers. Self-assembled monolayers were prepared by immersing gold-coated silicon wafers in a 0.5 mM solution of alkanethiol (dodecane thiol (98+% pure) or mercaptoundecanoic acid (95% pure)) in ethanol (200 proof, Ricca Chemical Co., Arlington, TX) for 48 h. Subsequently, samples were rinsed, sonicated for 5 s, and rinsed again in ethanol prior to drying under a nitrogen stream and storage in a nitrogen atmosphere. Sodium cationization was achieved by immersing the samples in a 10 mM aqueous solution of NaOH, for 1 min. Samples were then rinsed under a stream of ethanol for 10 s and dried under a nitrogen stream. Mercaptonicotinic Acid on Au. Gold-coated silicon wafers were immersed in a 10 mM solution of mercaptonicotinic acid (90% pure) in ethanol for 48 h. Subsequently, samples were rinsed, (22) Van Stipdonk, M. J.; English, R. D.; Schweikert, E. A. Anal. Chem. 2000, 72, 2618. (23) Karas, M.; Bahr, U.; Giebmann, U. Mass Spectrom. Rev. 1991, 10, 335.

3706

Langmuir, Vol. 20, No. 9, 2004

McArthur et al.

Figure 2. ToF-SIMS detection of the molecular ion of Angiotensin II from a mixture of the protein and the matrix dihydroxybenzoic acid (DHB) deposited onto gold. resolution C 1s and S 2p spectra were fitted using a Shirley background subtraction and a series of Gaussian peaks. Time-of-Flight Secondary Ion Mass Spectrometry. ToFSIMS analysis of all surfaces was performed with a PHI Model 7200 reflectron time-of-flight secondary ion mass spectrometer (Physical Electronics, Eden Prairie, MN) equipped with an 8 keV Cs+ primary ion source. The post-acceleration voltage was fixed at 10 kV throughout with positive ion spectra acquired by rastering the ion beam over a 100 µm × 100 µm sample area with data collected in the 0-1500 m/z range. The primary ion dose was less than 1012 ions/cm2 to maintain static ToF-SIMS conditions25 and charge neutralization was achieved where necessary with a pulsed electron flood gun. Spectra were calibrated to the CH3+, C2H3+, C3H5+, and C7H7+ peaks. The AuSCH2+ peak was included in the calbiration where appropriate. A minimum of five samples and three points per sample were analyzed for each sample type.

Results Figure 1. Sample preparation methods used to induce protein molecular ions in ToF-SIMS. (a) Matrix-assisted ToF-SIMS using dihydroxybenzoic acid (DHB), (b) sodium cationization of carboxyl-terminated alkane thiol assembled onto gold, and (c) self-assembly of mercaptonicotinic acid (MNA) onto gold sonicated for 5 s, and rinsed again in ethanol, prior to drying under a nitrogen stream and storage in a nitrogen atmosphere. Polymer and Protein Deposition. All proteins used in the study, Angiotensin II, substance P, and Ala-Pro-Gly-[Ile3, Val5]Angiotensin II, were dissolved in water at a final concentration of 0.1 and 0.01 mg/mL. A mixture of the three proteins was also prepared with the final concentration of each protein fixed at 0.1 mg/mL. A 0.5 mg/ mL solution of poly(ethylene glycol) (1000 MW) was prepared in methylene chloride. Samples were prepared by either spin coating or adsorbing the analyte onto the various substrates. Spin coating was performed by rotating the sample at 6000 rpm and then applying 5 µL of the protein or polymer solution and allowing the sample to spin for a further 20 s. Protein adsorption was performed by immersing the substrate in a vial with 2 mL of protein solution. Samples were incubated at 37 °C for 2 h, then rinsed thoroughly in water, and dried under a stream of nitrogen. X-ray Photoelectron Spectroscopy (XPS). XPS Analysis was performed using a SSI X-probe equipped with a monochromated Al KR source (1000-µm spot). Elements present on the surface were identified from a survey scan. For further analysis and quantification, spectra from the individual peaks were collected at 150 eV pass energy (res 4). High-resolution spectra were collected at 50 eV (res 2). All data were collected at a takeoff angle of 55° from the surface normal. A value of 285 eV for the binding energy of the main C 1s component (CHx) was used to correct for charging of specimens under irradiation.24 High(24) Beamson, G.; Briggs, D. High-resolution XPS of organic polymers. The Scienta ESCA300 database; John Wiley and Sons: Chichester, 1992.

Matrix-Enhanced SIMS (MESIMS). ToF-SIMS analysis of the samples prepared using the MALDI-MS matrixes dihydroxybenzoic acid (DHB) and sinapinic acid (SA) indicated that the addition of the matrix aided the generation of molecular ions from a range of small proteins < 2000 MW. Figure 2 illustrates a typical spectra produced when the matrix (DHB) and protein (Angiotensin II) were premixed and the sample applied to a gold (Au) substrate. Peaks were also occasionally detected with m/z corresponding to [M+Na]+ and [M+H]2+ ions. In each instance a series of peaks was detected following the protein ions. Analysis of the peak position and their areas indicated that they correlated with the isotopic pattern calculated for the protein molecular ion from the isotopic abundances of C, H, O, and N, rather than with a series of [M+nH]+ peaks. Sequences of peaks correlating with the isotopic abundances of the [M+H]2+ and the [M+Na]+ ions were also detected (data not shown). While a signal could be produced occasionally from samples applied to the silicon (Si) substrates, molecular ion peaks were generally below 20 counts and the signalto-noise ratio (s/n) in the higher mass regions (>1000 m/z) was poor. One of the greatest difficulties lay in reproducing signal across the sample area. Visual inspection of the samples using the instrument’s charge coupled detector television (CCTV) indicated that the matrix/protein crystals were unevenly distributed across the substrate and the three-dimensional nature of crystal surface made it difficult to optimize the analysis parameters. To investigate the effect of salt on the molecular ion yield, matrix solutions were passed through a cation (25) Marletta, G.; Catalano, S. M.; Pignataro, S. Surf. Interface Anal. 1990, 16, 407.

Protein Molecular Ions in ToF-SIMS

Langmuir, Vol. 20, No. 9, 2004 3707 Table 1. XPS Atomic Composition Data Following Each Stage in the Sample Preparation Process for the Cationized Self-assembled Monolayersa XPS atomic composition sample Au control Au control Au + Ads Angio II Au + Ads Angio II C10COOH C10COONa C10COONa + Ads Angio II C10COONa + Ads Angio II C10COONa + SC Angio II C10COONa + SC Angio II

atomic ratios

C 1s Au 4f O 1s N 1s S 2p Na 1s C:Au

O:C

38.9 41.8 50.4 51.6 61.3 53.4 57.7

0.68 0.77 1.61 1.69 2.88 2.25 2.94

0.10 0.10 0.24 0.23 0.23 0.29 0.24

57.0 54.1 31.3 30.4 21.3 23.8 19.6

4.1 4.1 11.9 6.5 11.6 6.4 14.0 15.6 13.8 7.5

2.8 2.5 1.4

58.2 22.8 12.8 4.4

1.8

2.55

0.22

58.4 24.0 11.7 4.1

1.9

2.44

0.20

58.0 26.4 11.8 1.7

2.0

2.20

0.20

4.7

a Ads ) protein adsorbed from solution; SC ) protein spun cast from solution.

Figure 3. ToF-SIMS detection of the molecular ion of Angiotensin II from a mixture of the protein and desalted DHB deposited onto (a) silicon and (b) gold.

exchange column. Samples were then prepared using this desalted matrix solution. Figure 3 illustrates some of the typical results seen with the deposition of Angiotensin II/DHB mixtures on Si and Au substrates. Visually the resulting sample crystals were a finer and more evenly distributed, and protein molecular ion signals could be detected at virtually any point within the sample with a range of proteins