Two-Step Matrix Application Technique To Improve Ionization

Anal. Chem. 2006, 78, 8227-8235

Two-Step Matrix Application Technique To Improve Ionization Efficiency for Matrix-Assisted Laser Desorption/Ionization in Imaging Mass Spectrometry Yuki Sugiura,† Shuichi Shimma,‡,§ and Mitsutoshi Setou*,†,‡,§

Department of Bioscience and Biotechnology, Tokyo Institute of Technology, Yokohama, Kanagawa 226-8501, Japan, School of Life Science, Department of Physiological Sciences, The Graduate University for Advanced Studies, Okazaki, Aichi 444-8585, Japan, and Okazaki Institute for Integrative Bioscience, National Institutes of Natural Sciences, Okazaki, Aichi 444-8787, Japan

A novel matrix application protocol for direct tissue mass spectrometry is presented. Matrix-assisted laser desorption/ionization is a popular ionization procedure for direct tissue analysis and imaging mass spectrometry. Usually, matrixes are applied by dispensing droplets through either pipettes or automated dispensing machines, or by airbrushing. These techniques are very simple, but it was difficult to obtain uniform matrix crystals on the tissue surface, and nonuniform crystals degrade the spectrum qualities. Here we report a new matrix application protocol, which is a combination of spraying and dispensing droplets, and we have succeeded in overcoming these problems in conventional matrix applications on tissue surfaces. We call our new technique the “spray-droplet method”. In this technique, tiny matrix crystals formed by spraying act as seeds for crystal growth. Our technique leads to matrix spots that are filled homogeneously with minute crystals. Such matrix crystals dramatically improve peak intensity and signal-to-noise ratio. In an example on a rat brain section, the number of detectable peaks was increased and signal intensity of m/z 5440 in our method was ∼30.6 times higher than that in conventional methods. We used this spray-droplet method with a chemical ink-jet technology for matrix deposition to succeed in MALDI imaging of signals, which were undetectable from the conventional matrix applications. The intact tissue analysis of biological compounds using mass spectrometry (MS) is a subject of much interest for the next generation of MS. Mass spectra from tissue slices are used to obtain a molecular pattern in biological samples. Recent studies have used this new technique successfully in molecular imaging for biological applications,1-5 pathological applications,6-14 and drug discovery.15,16 * Corresponding author: (e-mail) [email protected]; (fax) +81-564-59-5291; (tel) +81-564-59-5267. † Tokyo Institute of Technology. ‡ The Graduate University for Advanced Studies. § Okazaki Institute for Integrative Bioscience. 10.1021/ac060974v CCC: $33.50 Published on Web 11/11/2006

© 2006 American Chemical Society

To carry out successful direct tissue analysis with matrixassisted laser desorption/ionization (MALDI), samples must be prepared by proper techniques.17,18 In the sample preparation procedures generally used today, tissue sections are usually washed with 70% ethanol for 30 s twice to wash the salts away.18 Then a matrix solution is dispensed by hand or by an automated dispenser19-22 directly onto the tissue surface. However, there are disadvantages in the direct-dispensing procedure (i.e., directdroplet method) in which the droplets of matrix solution diffuse on the tissue surface and results in the nonuniform matrix crystals that form in the droplet area. Spraying matrix solution by an airbrush has also been used for MALDI imaging.7 This technique can form very tiny matrix crystals on the surface, but the intensity of the signals is not satisfactory compared to the droplet method. Recently, a technique for matrix seeding on a tissue surface was reported.19 In this method, ground matrix powder was distributed using a paintbrush over the sample surface.19 According to this study, the powder distribution method improved crystal coverage and signal uniformity. A few other strategies improving mass resolution and sensitivity from tissues were also reported, (1) Chaurand, P.; Schwartz, S. A.; Reyzer, M. L.; Caprioli, R. M. Toxicol. Pathol. 2005, 33, 92-101. (2) Caldwell, R. L.; Caprioli, R. M. Mol. Cell. Proteomics 2005, 4, 394-401. (3) Chaurand, P.; Schwartz, S. A.; Caprioli, R. M. Curr. Opin. Chem. Biol. 2002, 6, 676-681. (4) Stoeckli, M.; Chaurand, P.; Hallahan, D. E.; Caprioli, R. M. Nat. Med. 2001, 7, 493-496. (5) Caprioli, R. M.; Farmer, T. B.; Gile, J. Anal. Chem. 1997, 69, 4751-4760. (6) Rohner, T. C.; Staab, D.; Stoeckli, M. Mech. Ageing Dev. 2005, 126, 177185. (7) Schwartz, S. A.; Weil, R. J.; Johnson, M. D.; Toms, S. A.; Caprioli, R. M. Clin. Cancer Res. 2004, 10, 981-987. (8) Pierson, J.; Norris, J. L.; Aerni, H. R.; Svenningsson, P.; Caprioli, R. M.; Andren, P. E. J. Proteome Res. 2004, 3, 289-295. (9) Yanagisawa, K.; Xu, B. J.; Carbone, D. P.; Caprioli, R. M. Clin. Lung Cancer 2003, 5, 113-118. (10) Chaurand, P.; Fouchecourt, S.; DaGue, B. B.; Xu, B. J.; Reyzer, M. L.; Orgebin-Crist, M. C.; Caprioli, R. M. Proteomics 2003, 3, 2221-2239. (11) Yanagisawa, K.; Shyr, Y.; Xu, B. J.; Massion, P. P.; Larsen, P. H.; White, B. C.; Roberts, J. R.; Edgerton, M.; Gonzalez, A.; Nadaf, S.; Moore, J. H.; Caprioli, R. M.; Carbone, D. P. Lancet 2003, 362, 433-439. (12) Stoeckli, M.; Staab, D.; Staufenbiel, M.; Wiederhold, K. H.; Signor, L. Anal. Biochem. 2002, 311, 33-39.

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006 8227

such as the deposition of a thin layer of gold on the tissue.23,24 However, there is much room for further optimization and innovation. In this paper, we describe a matrix seeding procedure to improve not only the signal uniformity but also the sensitivity and signal-to-noise ratio in direct tissue mass spectrometry. Our technique consists of two steps: spray coating of a low-concentration matrix solution and depositing of a higher-concentrated matrix solution. We have designated this procedure “the spray-droplet technique”. Our experimental results show that this method improves not only crystal density and uniformity but also crystal fineness in the matrix spot, thus improving sensitivity and signal-to-noise ratio. Furthermore, spectrum improvements resulted in the increase of the number of signals that could be visualized on the tissue section by MALDI imaging experiment. The keys to this technique are an iterative spray-coating and a controlled humidity. Finally, this paper presents the improvement of the spectrum and MALDI imaging results by our spray-droplet method. EXPERIMENTAL SECTION Chemicals. Acetonitrile (ACN) and trifluoroacetic acid (TFA) were purchased from Kanto Chemical (Tokyo, Japan). Octyl β-Dglucopyranoside (n-octyl glucoside) was obtained from Sigma (St. Louis, MO). Calibration standard protein (5000-17 000 Da) and sinapinic acid (SA) were purchased from Bruker Daltonics (Leipzig, Germany). Mass Spectrometry. Mass spectrometry was performed with a MALDI-TOF/TOF-type instrument, Ultraflex 2 TOF/TOF (Bruker Daltonics). The instrument was equipped with a 355-nm Nd: YAG laser. Data were acquired in the positive-ion mode using an external calibration method. The external calibration protein was deposited on the surfaces of sample support materials to minimize mass shift. The laser irradiated the tissue surface with 100 shots. The spectra shown in the Results and Discussion section were accumulated during 500 consecutive laser shots. Section Preparation. This study used a cerebrum of 8-weekold male C57BL/6Cr mice and a liver and a cerebellum of a female (13) Masumori, N.; Thomas, T. Z.; Chaurand, P.; Case, T.; Paul, M.; Kasper, S.; Caprioli, R. M.; Tsukamoto, T.; Shappell, S. B.; Matusik, R. J. Cancer Res. 2001, 61, 2239-2249. (14) Chaurand, P.; DaGue, B. B.; Pearsall, R. S.; Threadgill, D. W.; Caprioli, R. M. Proteomics 2001, 1, 1320-1326. (15) Rudin, M.; Rausch, M.; Stoeckli, M. Mol. Imaging Biol. 2005, 7, 5-13. (16) Rudin, M.; Allegrini, P.; Beckmann, N.; Gremlich, H. U.; Kneuer, R.; Laurent, D.; Rausch M.; Stoeckli, M. Ernst Schering Res. Found. Workshop 2004, 47-75. (17) Sugiura, Y.; Shimma, S.; Setou, M. J. Mass Spectrom. Soc. Jpn. 2006, 54, 45-48. (18) Schwartz, S. A.; Reyzer, M. L.; Caprioli, R. M. J. Mass Spectrom. 2003, 38, 699-708. (19) Aerni, H. R.; Cornett, D. S.; Caprioli, R. M. Anal. Chem. 2006, 78, 827834. (20) Aerni, H. R.; Erskine, A. R.; Reyzer, M. L.; Lee, D.; Cornett, D. S.; Caprioli, R. M. Proc. Am. Soc. Mass Spectrom. 51st Ann. Conf. Mass Spectrom. Allied Top.; Montreal, Canada; 2003. (21) Nakanishi, T.; Ohtsu, I.; Furuta, M.; Ando, E.; Nishimura, O. J. Proteome Res. 2005, 4, 743-747. (22) Ohtsu, I.; Nakanisi, T.; Furuta, M.; Ando, E.; Nishimura, O. J. Proteome Res. 2005, 4, 1391-1396. (23) Altelaar, A. F.; Klinkert, I.; Jalink, K.; de Lange, R. P.; Adan, R. A.; Heeren, R. M.; Piersma, S. R. Anal. Chem. 2006, 78, 734-742. (24) Lemaire, R.; Tabet, J. C.; Ducoroy, P.; Hendra, J. B.; Salzet, M.; Fournier, I. Anal. Chem. 2006, 78, 809-819.

8228

Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

Wistar rat (Crl:WU). Animals were sacrificed and dissected under diethyl ether anesthesia. The extirpated tissue blocks were immediately frozen in powdered dry ice and stored at -80 °C until needed. Frozen sections were sliced at -16 °C with a cryostat (Leica CM 3050) at a thickness of 5 µm.17 To fix each tissue block, an optimum cutting temperature (OCT) polymer was used. When the sections were sliced, the cutting block was not embedded in OCT, since any residual polymer on the tissue slices might have degraded the mass spectra.18 Frozen sections were thaw-mounted on indium-tin oxide-coated sheet (Tobi Co., Ltd., Kyoto, Japan) and allowed to dehydrate in a vacuum chamber for 30 min. The dried tissues were then rinsed with 70% ethanol for 30 s twice and then were dehydrated for 15 min. Before the matrix was applied, 1 mL of 20 mM n-octyl glucoside solution25 was uniformly sprayed over the tissue surface. All solutions were sprayed by a 0.2-mm nozzle caliber airbrush, Procon Boy FWA Platinum (Mr. Hobby, Tokyo, Japan). Protocol for the Spray-Droplet Matrix Application Method. In initial experiments, we confirmed the matrix seeding effect of the spray-droplet method, which enhanced homogeneous crystallization in a matrix spot on the metal-coated glass surface. However, in the case of the biological tissue surface, which has a highly complicated structure compared with metal-coated material, control of crystal growth conditions was essential for highly efficient ionization. Day-to-day variability of such conditions can result in poor reproducibility.19 Our experiment revealed that humidity and temperature were key factors in the droplet drying time. We found that spraying under a humid condition leads to successful micronuclei formation on the tissue surface with a good reproducibility. Further experiments proved that an iterative spray-coating and the control of humidity improves crystal fineness and density as well as the homogeneous distribution in the matrix spot, as discussed below. As a result of our protocol, these iterations resulted in sufficient seeding effects. In contrast, spray-coating performed under low humidity kept by silica gel did not lead to sufficient seeding effects reproducibly. The microcrystal nuclei formation on the tissue surface was not observed under the stereomicroscope. We considered that crystallization occurring between the airbrush’s nozzle tip and the tissue surface induced this phenomenon. The tiny crystallized matrix material may be resolved by the wetness of the tissue section, and such a process may have made the material amorphous. Matrix solutions were prepared at two concentrations: 4.0 mg/ mL for spraying and 10 mg/mL to produce droplets. Both solutions contained 50% ACN diluted with 0.1% TFA. Matrix concentrations are important because a solution that is too highly concentrated for spraying (>4.0 mg/mL) leads to excess aggregation of crystals on the tissue surface. For the dispensing solution, a diluted concentration (e10 mg/mL) leads to greater numbers of minute crystals in the matrix spot.26 The low-concentration matrix solution (4.0 mg/mL) was sprayed with the airbrush. In this study, the distance between the brush’s nozzle tip and the tissue surface was kept at 15 cm. In the following explanation, the spraying period was fixed at 30 s. (25) Katayama, H.; Nagasu, T.; Oda, Y. Rapid Commun. Mass Spectrom. 2001, 15, 1416-1421. (26) Sadeghi, M.; Vertes, A. Appl. Surf. Sci. 1998, 129, 226-234.

Figure 1. Schematic diagram of the spray-droplet procedure. (a) Thawed tissue section on the supporting material. (b) Spray-coating under humid environment. (c) Depositing droplet on the tissue section.

To improve the spectrum quality for direct tissue analysis, spray-coating must be performed in a humid environment, as follows: the tissue sections on the supporting material were placed in a humidified box saturated with Milli-Q water steam, and the box was opened only during spray coating. The box was incubated at 37 °C before use. The tissue surfaces became slightly moist as a result of the 30-s spraying. We applied a 5-min interval between sprays; during this interval, most of the tissue surface dried. This operation was repeated four times. Then the matrix solution for droplets was deposited manually in the spot profiling. On the other hand, we used a chemical ink-jet printer (CHIP-1000, Shimadzu, Kyoto, Japan) to make arrays of matrix spots for MALDI imaging. As previous works pointed out,19 it was difficult to maintain stable ejection of concentrated matrix solution (>10 mg/mL) because of the clogging of the matrixes at the nozzle tip when we used an automatic dispenser. We optimized the composition of matrix solution and overcame the problem. The concentration of matrix solution was 8 mg/mL in 40% ACN containing 0.1% TFA. After spraying the low-concentration matrix solution, a total amount of 50 nL of matrix solution was printed onto the tissue surface. The solution was printed as 50 iterations onto the surface at 1 nL/ iteration. The interval between each spot was set at 300 µm. Real-Time Observation of Crystal Formation. The crystallization processes between the spray-droplet and direct-droplet method were compared. The observation was performed under a stereoscopic microscope (SZX7, Olympus, Tokyo, Japan). Each crystallization process was recorded with a digital camera (Camedia C-5060WZ, Olympus), and the microscope and camera were connected via an optical adaptor (NY2000S3 Super Adaptor, Olympus). Time-lapse evolution pictures in the Results and Discussion section were edited with QuickTime Pro. Scanning Electron Microscope Observation. The matrix spots on the tissue sections were coated with a platinumpalladium compound in a magnetron ion sputter coater (E1030; Hitachi, Tokyo, Japan) and were observed with a Hitachi S-4500 field emission-type scanning electron microscope (SEM). All images were obtained using the lower detection mode of the SEM instrument. Quantitative Measurements. To prove the uniformity of the crystals formed in the spots, we performed mass imaging of the matrix spots formed by the spray-droplet and direct-droplet methods. The matrix solution was deposited onto the liver section. Each matrix spot was scanned using laser spots of 60 µm on a 100-µm raster. For all signals derived from proteins in the liver section, we compared the differences in ion signal distribution in the spots, and evaluated the standard deviation (SD) of signal intensities.

MALDI Imaging of Adult Rat Cerebellum Sections. Adult rat cerebellum sections were prepared with three different matrix application procedures, which were the spray-droplet method, direct-droplet method, and spray-coating. The raster scan on the tissue surface was performed automatically. The number of laser irradiations was 100 shots in each spot. The interval of data points was 100 µm, giving a total of ∼1200 data points for each tissue section. In order to assemble the obtained spectra to ion image and correct the spectra with smoothing and baseline subtraction, data analysis software (Flex imaging ver. 1.0 and FlexAnalysis ver. 2.4) were used. RESULTS AND DISCUSSION Characteristics of Matrix Spots Formed by the SprayDroplet Method. Figure 1 is a schematic diagram of our twostep matrix spray-droplet application method: spray-coating of dilute matrix solution and depositing of matrix solution. Matrix deposition was performed with automatic dispenser or manual micropipet. Figure 2 shows time-lapse observation of the crystallization process in the matrix droplets. To characterize the matrix spots formed by our method, we deposited a small amount of matrix solution on the spray-coated surface of the mouse brain section (Figure 2a-d) and on the noncoated surface, using a micropipet (Figure 2e-h). The matrix solution applied to the spray-coated tissue surface resulted in coverage of the matrix spot densely packed with an enormous number of microcrystals homogeneously (Figure 2d). In contrast, the same volume of solution applied to a noncoated surface formed larger crystals distributed discretely (Figure 2h). We found that tiny spray-coated matrix crystals worked as seeds for crystal growth. Immediate crystallization was another feature of the spraydroplet method. On the spray-coated tissue surface, it took 60 s for ∼0.1 µL of matrix solution to dry up, while the same volume solution required 90 s on the noncoated surface. We interpret this phenomenon to mean that crystals in the solution can start growing with existing crystal nuclei. Moreover, crystallization occurred homogeneously over the whole droplet area. Such rapid crystallization induced the depression of the boiling point in the matrix solution, thus leading to the immediate evaporation of solvent. This feature prevented local and large crystal growth, which can cause analyte migration across the droplet area. In contrast, migration of needlelike crystals and local crystal growth were observed in the matrix spot formed by the direct-droplet method. In the matrix spot, crystal nucleation occurred in the applied droplet, after which crystallized matrix materials migrated Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

8229

Figure 2. Time-lapse observation of crystal formation on spray-coated surface (a-d) and without coated surface (e-h). Crystals observed with scanning electron microscope in spray-droplet (i) and direct-droplet matrix spot (j).

across the droplet area. Then, they coalesced into larger crystals. As a result, only unfavorably large needlelike crystals were formed. Inhomogeneous distribution of crystals and analyte migration across the droplet area may perturb the accuracy of the signal distribution, which is sometimes problematic in mass imaging. We considered that the SEM observation of crystal size on the tissue surface is highly useful for a better comprehension of the proposed method’s contribution to spectrum improvement in direct tissue analysis.27 We observed the crystals using SEM instruments. Figure 2i shows SEM images of crystals obtained from the spray-droplet method and Figure 2j was from the direct-droplet method. Polygonal columnar crystals, typically hexagonal columns, were found in sizes of 2-30 µm (Figure 2i). The microcrystal on the tissue surface was as small as that obtained by the traditional proteomics improvement.28 On the other hand, far fewer crystals were formed in the matrix spot formed by the direct-droplet method. The crystals were much larger and complex, with fused hollow columns (Figure 2j). Probably the columns grew slowly from one nucleus into a radial pattern column. The principal differences between the spray-droplet and direct-droplet methods were the high density and fineness of the microcrystals. (27) Vaidyanathan, S.; Winder, C. L.; Wade, S. C.; Kell, D. B.; Goodacre, R. Rapid Commun. Mass Spectrom. 2002, 16, 1276-1286. (28) Onnerfjord, P.; Ekstrom, S.; Bergquist, J.; Nilsson, J.; Laurell, T.; MarkoVarga, G. Rapid Commun. Mass Spectrom. 1999, 13, 315-322.

8230


Uniformity of Ion Signal Distribution Assessed by MALDI Imaging. Figure 3 shows a comparison of the matrix crystal uniformity between spray-droplet and direct-droplet methods by one spot MALDI imaging on the liver section. As examples, 12 consecutive spectra from a novel spot and a conventional spot are shown in Figure 3a and b, respectively. The data points were on the white and dashed lines, represented in the optical images of each spot (Figure 3c and g). The variances in signal intensity among the shots were drastically improved in the spray-droplet spot. Imaging results of the matrix spot suggest that the spraydroplet spot produces more homogeneous ion signals than the direct-droplet spot. This result indicates that uniform spectrum quality was acquired from almost all data points with the same laser condition. Since nonuniform crystal perturbs the accuracy of the signal distribution of the tissue section, our proposed procedure overcame the problem in imaging mass spectrometry. Statistical analysis shows that our method exhibited only 15-40% of the SD of signal intensity compared to the direct-droplet method (Figure 3f). We can conclude that the spray-droplet method produces a more homogeneous distribution of signals over the matrix spot by uniform crystallization. Spray-Droplet Method Improved Spectrum Quality in MALDI-MS. The quality of the spectrum obtained from the spraydroplet spot (Figure 4a) was drastically improved in terms of higher intensity and signal-to-noise ratio relative to the spectrum

Figure 3. Evaluation of crystal uniformity. Mass spectra obtained from line scan in spray-droplet (a) and direct-droplet matrix spot (b). The data points are on the white and dashed lines shown in (c) spray droplet and (d) direct droplet. MALDI imaging result for each spot: (d-f) spray droplet and (i-k) direct droplet. Scale bar, 1.0 mm. Comparison of standard deviation in obtained data sets (f).

from the direct-droplet spot (Figure 4b). Figure 4c is the enlarged spectrum of Figure 4b; the vertical axis was normalized to 3% of the original axis. In this comparison, the mouse brains were cut into coronal sections with bilateral symmetry. One hemisphere was spray-coated while the other was screened out. Then the same volume of matrix solution was applied to the cerebral cortex

section of both sites in a line-symmetric way. We detected more intense signals from the microcrystal spot; for example, 30.6 times higher for m/z 5440 and 29.7 times higher for m/z 6569 than those acquired from the direct-droplet spot (Figure 4a). On average, 15 major signals commonly found in the two spectra with high intensity (Figure 4a and c), which were presumed to be derived Analytical Chemistry, Vol. 78, No. 24, December 15, 2006

8231

Figure 4. Comparison of mass spectra from mouse brain section. (a) spray droplet spot and (b) direct droplet. (c) Enlarged spectrum of (b).

from abundant proteins, had 32.3 times higher intensity in the spray-droplet spot. Moreover, the signal-to-noise ratio was dramatically improved, such as a 9.4-fold increase for m/z 5440 and a 9.1-fold increase for m/z 6569. On average, for the 15 signals described above, a 9.9-fold higher signal-to-noise ratio was recorded in the spray-droplet spot. As a consequence of such improvement, ∼290 signals were detected in Figure 4a, giving ∼90 more peaks than those detected when the direct-droplet method was used. This spectrum improvement using the spray-droplet method was accounted for as follows; we demonstrated that the principal differences of crystal morphology between the spray-droplet and direct-droplet matrix spot were the high density and fineness of the microcrystals (Figure 2i and j). These two differences are the most important factors in spectrum improvement. The morphology of crystal is known to influence spectrum quality in MALDI-MS of protein profiling using SA in the traditional proteomics procedure.26 The crystal form is influenced by sample preparation strategies, i.e., the protein/matrix cocrystallization conditions.26-29 Sample preparation strategies have a particularly strong influence on the distribution of crystal size.26,28 It has been reported that smaller crystals at high density on a MALDI target plate increase volatilization by laser irradiation.26 Moreover, in the traditional proteomics procedure, sample preparation methods that lead to the formation of smaller crystals have been shown to produce better reproducibility,30 improved resolution, higher sensitivity,31 and adequate signal-to-noise-ratios in fewer laser shots.32 (29) Beavis, R. C.; Bridson, J. N. J. Phys. D Appl. Phys. 1993, 26, 442. (30) Chaurand, P.; Schwartz, S. A.; Billheimer, D.; Xu, B. J.; Crecelius, A.; Caprioli, R. M. Anal. Chem. 2004, 76, 1145-1155. (31) Vorm, O.; Roepstorff, P.; Mann, M. Anal. Chem. 1994, 66, 3281-3287. (32) Einberger, S. R.; Boernsen, K. O.; Finch, J. W.; Robertson, V.; Musselman, B. E. Proc. Am. Soc. Mass Spectrom. 41st Ann. Conf. Mass Spectrom. Allied Top.; San Fransisco, CA, 1993.

8232


Another factor that contributed to this improvement was enhanced analyte extraction in the spray-drop method. Under the humidified condition, the tissue slices became slightly moist by the spray coating of matrix solution, and the tissue surface dried up instantly. This phenomenon means that sprayed solution forms very tiny droplets on the tissue surface and evaporates immediately. More proteins can be incorporated into the growing crystal from scattered nuclei by selective interaction with one of the crystal faces.29 In addition, in direct tissue analysis, it has been reported that iterative dispensing of matrix solution at the same coordinates led to higher quality mass spectra.20 Besides this fact, our results suggested that the iterative matrix application contributed to the extraction of analytes from inside tissue. A sufficient quantity of analytes contributed to this spectrum improvement. Evaluation of Spectrum in MALDI Imaging Mass Spectrometry. Figure 5 shows evaluation of spectrum quality, which was obtained with the spray-droplet and direct-droplet methods employing an automated dispenser and a spray-coating procedure in MALDI imaging experiment. Three successive tissue sections were prepared for an imaging experiment employing spray-droplet method (Figure 5a), direct-droplet method (Figure 5b), and spraycoating (Figure 5c). In Figure 5a and b, matrix solution was printed as quadrate array with a total of 230 spots and 300 µm spot-to-spot interval. A total of 50 nL/spot of matrix solution was dispensed and the average diameter of spots became 200 µm. The 50 nL of matrix solution was printed as 50 iterations at 1 nL/iteration. Automatic data acquisition was performed and resulted in ∼1200 spectra for each method. After data acquisition, we evaluated the spectrum quality obtained by each technique. From the total spectra obtained, 16 spectra from two different regions of a rat brain section were accumulated, which are cerebellar cortex region and medulla of cerebellum region (represented as white squares A and B, respectively, in Figure 5a-c). Accumulated mass spectra were shown in Figure 5d and e. From the present spectra, it is clear that spray-droplet spots generated the highest intensity and signal-to-noise ratio for almost all detected signals as we have shown in single spot analysis. Signals marked with an asterisk were detected only in the mass spectra obtained from spray-droplet spots. Figure 5f shows the number of detected signals using each matrix application method, and it indicates improved signal detection efficiency of spraydroplet method. Emphasis is placed on that by using the spraydroplet method; these unique signals were detected as major signals in the spectrum, for example, m/z 4506 from cerebellar cortex region and m/z 4999 from medulla of cerebellum. In the MALDI imaging experiment, such improvement of spectrum quality using the spray-droplet method in terms of higher signal-to-noise ratio and signal intensity can provide a rise in detectable signals. An increase in total ion amount enables visualization of low-intensity signals and higher signal resolution, which discriminates similar molecular weight signals. Additionally, in the imaging experiment, several hundreds of matrix spots were formed on the tissue surface and it is useful to assess the reproducibility and reliability of spray-droplet method. Application of the Spray-Droplet Method to MALDI Imaging. Figure 6 shows examples of the MALDI imaging results using

Figure 5. Comparison of mass spectra obtained from an adult rat brain section. Optical images of microspotted tissue sections employing spray-droplet (a), direct-droplet (b), and spraycoating (c) method. Scale bar, 1.0 mm. Accumulated mass spectra collected from different region of rat brain section are shown (d, e). These regions are represented as white squares in Figure 6, cerebellar cortex (d) and medulla of cerebellum region (e). In each spectrum, asterisks represent major unique signals for spectra using the spray-droplet method. Number of detected signals in mass range of 2000 < m/z < 30000 from each region are shown in (f).

the spray-droplet method (a and d), direct-droplet method (6 and e), and spray-coating (c and f). The most important advantage of the spray-droplet method is the imaging capability of protein signals, which were undetectable by the other two methods. This is the great advantage in the tissue analysis, e.g., in biomarker discovery. Panels a-d in Figure 6 show that only the spray-droplet method can reconstruct the image of m/z 4999, which is strongly distributed in cortex region and m/z 4506 in medulla region. The other two methods failed to obtain a significant image. Spray-droplet and direct-droplet methods resulted in images with a similar lattice pattern that derived from discrete matrix

spots (Figure 6a, b, d, e). From these images, it was confirmed that seed crystals formed by spraying did not produce a measurable ion signal. Using software that fills the gap of the spots, these data can convert to consecutive images.33 Panels d and f in Figure 6 are control imaging results of signals at m/z 14109 and 9970, which were commonly detectable using these three matrix application methods. These protein signals are detected as a major signal in each spectrum and speculated to be derived from abundant proteins. Each imaging result clearly revealed a strong distribution of the signal at m/z 14109 in the (33) Clerens, S.; Ceuppens, R.; Arckens, L. Rapid Commun. Mass Spectrom. 2006, 20, 3061-3066.


8233

Figure 6. Comparison of protein imaging in rat cerebellum sections. MALDI imaging obtained from spray-droplet method (a) and (d), directdroplet method (b) and (e), and spray-coating method (c) and (f).

cerebellar cortex region and the signal at m/z 9970 in the medulla of cerebelluml i.e., these three methods were all efficacious in visualizing the distribution of such major signals. In these images, the spray-droplet method visualized more accurate signal distribution with enhanced contrast (Figure 6d) than the direct-droplet method (Figure 6e). We have shown in single spot analysis that one consequence of the spray-droplet 8234


method is homogeneous distribution of crystals, which limits analyte migration in the droplet area. Present imaging results demonstrate this feature contributed to improve the quality of MALDI imaging. Finally, these imaging results demonstrated the reproducibility and reliability of improvement using our spray-droplet protocol proposed in this report.

CONCLUSIONS A new matrix application method, consisting of the spraying and the dispensing of dilute matrix solution, was presented. Our method improved signal intensity and signal-to-noise ratio while providing homogeneous signal distribution in the matrix spot. As a consequence, the number of detected peaks was increased. These features offer improvements not only for protein profiling in single matrix spots but also for imaging mass spectrometry. More signals can be detected with accurate signal distribution on the tissue section. Using our matrix application, we succeeded in MALDI imaging not only major proteins but also minor proteins that was not detected with direct-droplet procedures. This

technique will increase the potential of direct tissue analysis by high-efficiency ionization of biological compounds. ACKNOWLEDGMENT The authors thank Dr. Michio Sato and Ms. Yoshimi Hinohara (Mitsubishi Kagaku Institute of Life Sciences) for support and fruitful discussions on SEM. This work was supported by a Grantin-Aid under the SENTAN program of the Japan Science and Technology Agency, to M.S. Y.S. and S.S. contributed equally to this work. Received for review May, 26, 2006. Accepted September 28, 2006. AC060974V


8235

Two-Step Matrix Application Technique To Improve Ionization

Recommend Documents