Molecular Proteomics Imaging of Tumor Interfaces by Mass

Molecular Proteomics Imaging of Tumor Interfaces by Mass Spectrometry Suki Kang,⊥ Hyo Sup Shim,⊥ Jong Sik Lee,† Dong Su Kim,‡ Hak Yong Kim,§ Seong Hyun Hong,§ Pan Soo Kim,§ Joo Heon Yoon,| and Nam Hoon Cho*,⊥,|,∇ Department of Pathology, Yonsei University College of Medicine, Seoul, Korea, Bruker BioSciences Korea Company, Ltd., Daltonics Division, Seoul, Korea, Genomine Research Division, Genomine, Inc., Pohang Technopark, Pohang, Korea, Gyeonggi Bio-Center, Instrument Support Team, Suwon, Korea, Research Center for Human Natural Defense System, Yonsei University College of Medicine, Seoul, Korea, and Brain Korea 21 Project for Medical Science, Seoul, Korea Received July 28, 2009

Abstract: The specific molecular profiles of ovarian cancer interface zones (IZ), the region between tumors and normal tissues, were evaluated using a new method involving matrix-assisted laser desorption/ionization (MALDI)-imaging mass spectrometry (IMS). We analyzed three ovarian serous carcinomas using MALDI-IMS. Principal component analysis (PCA) was used to evaluate the quality of tissue spatial features based on MALDI-IMS, and for analysis of large data sets of MALDI-IMS. Twodimensional gel electrophoresis and fluorescence microscopy were used to verify interface-specific proteins. Unique profiles were identified for the tumors, the normal zone, and the IZ. Through MALDI analysis, two interfacespecific proteins, plastin 2 and peroxiredoxin 1 (PRDX 1), were identified as differentially regulated between zones. Fluorescence microscopy revealed high expression levels of plastin 2 and PRDX 1 along the IZ of ovarian tumors. This comparative proteomics study using tissue MALDIIMS suggested that the IZ is different from the adjacent tumor and normal zones, and that plastin 2 and PRDX 1 may be interface markers specific to ovarian tumors. Keywords: MALDI-IMS • interface • 2-dimensional gel electrophoresis • ovarian cancer

Introduction Ovarian cancer is one of the most lethal neoplasms in women, and serous carcinoma is the most common type. However, the molecular events that underlie the development of ovarian serous carcinoma are largely unknown.1 The identification of major molecular markers for ovarian cancer is challenging due to available sampling sizes and sampling bias. Additionally, the tendency toward cystic and hemorrhagic or * To whom correspondence should be addressed. Nam Hoon Cho, M.D. Dept. of Pathology, Yonsei University College of Medicine, Seoul, Korea. ⊥ Department of Pathology, Yonsei University College of Medicine. † Bruker BioSciences Korea Co., Ltd. ‡ Genomine, Inc. § Gyeonggi Bio-Center. | Research Center for Human Natural Defense System, Yonsei University College of Medicine. ∇ Brain Korea 21 Project for Medical Science. 10.1021/pr900666q

 2010 American Chemical Society

necrotic tumor formation, intratumoral heterogeneity, and the broad spectrum of biological behaviors of ovarian cancer, especially borderline malignancies, makes it more difficult to match gross lesions with microscopic findings. Prognostic and biomarker studies are likely more efficient within tumor interface zones (IZ), where active cancer invasion or epithelialmesenchymal transition (EMT) occurs. Molecular studies using state-of-the-art technology based on high-throughput screenings may be occasionally misleading without the support of pathologists. Since the late 1980s, matrix-assisted laser desorption/ionization (MALDI)-imaging mass spectrometry (IMS) has been used for detection and verification of peptides or other polymers of biological interest.3-5 MALDI-MS is a powerful analytical tool for screening for biomarkers of early cancer and indicators for prognostication, but it inevitably requires tissue lysates. However, MALDI-IMS is an emerging methodology that has evolved from MALDI tissue profiling, which enables the identification of large proteins form biological samples, by combining molecular analysis with examination of histological sections. Therefore, direct correlations between molecular analysis and tissue localization can be made.6-9 Direct application of MALDI-timeof-flight (TOF)-mass spectrometry (MS) on tissue sections makes it possible to obtain specific information on local molecular composition, relative abundance, and spatial distribution for image profiling.10-12 Despite the potential of MALDI-IMS, researchers should be cautious that output may be very limited. Common obstacles for acquisition of quality results include the quality of the tissue samples, difficult procedures for application of the matrix, and enzymatic digestion.13,14 We acquired a portrait of IZs in ovarian cancer by using MALDI-IMS, and verified IZ-specific proteins through 2-D gel electrophoresis and fluorescence microscopy.

Material and Methods Sample Preparation. Frozen samples of three ovarian serous cancer tissues were obtained from Yonsei University College of Medicine. The study was approved by the Hospital Ethics Committee(4-2008-0380). All samples were prepared with the consent of the tissue donors. Tissue samples were obtained immediately after surgery and stored at -80 °C. For histological Journal of Proteome Research 2010, 9, 1157–1164 1157 Published on Web 10/12/2009

technical notes processing, verification of ovarian cancer, and following MALDIIMS analysis, cryosections of the tissues were stained using hematoxylin and eosin (H&E).15 MALDI-IMS Analysis. MALDI-IMS. Cryosections of samples were produced (10 µm thickness) at -25 °C using a cryostat. The sections were transferred to precooled (-20 °C) conductive indium-tin-oxide (ITO)-coated glass slides. Glass slides and tissues were thawed together. Sections were washed twice for 15-30 s in 70% ethanol and once for 15 s in 100% ethanol. Sections were air-dried and stored at -80 °C. The MALDI matrix was applied using an ImagePrep station (Bruker Daltonics, Bremen, Germany) with no pretreated digestion as described in the manufacturer’s protocol. The matrix for MALDI measurement was 10 mg/mL sinapinic acid (SA) in 50% acetonitrile with 0.2% trifluoroacetic acid (TFA). MALDI measurements and image analyses were performed using a linear Autoflex instrument equipped with a Smartbeam laser and FlexImaging 2.1 and ClinProTools 2.1 software packages (Bruker Daltonics). MALDI measurements were done in linear mode in a mass range of 3000-30 000 Da with a sampling rate of 0.1 GS/s.15 Spectra were accumulated during 100 consecutive laser shots. The interval between data points was 80-200 µm. FlexImaging 2.0 (Bruker Daltonics) was used for the reconstruction of images from the spectra. The image files were constructed from a data set obtained from matrix spots of 80-200 µm in diameter. The lateral resolution of the images produced in this process depended on the size of the laser ablation area. Spatial resolution was set to 80-200 µm.16 Following MALDI-IMS analysis, the sections were stained with H&E, scanned at 1200 dpi resolution, and coregistered with MALDI-IMS results. PCA Statistical Analysis. Statistical analyses were carried out using ClinProTools 2.1 software (Bruker Daltonics). Mass spectra were internally recalibrated on common spectral alignment peaks, and normalized to the total ion count. An average spectrum created from all single spectra was used for peak picking. Signal intensities were used for all calculations. For PCA, individual peak intensities were standardized across the data set. For PCA analysis, mass spectra were selected for the tissue by assigning representative areas of the interface, as well as the tumor and normal zones. For statistical calculations, one average spectrum was generated per section.17,18 Two-dimensional (2-D) Gel Electrophoresis, Image Analysis, and Protein Identification by MS. Protein Extraction. For tissue protein profiling, frozen tissue was directly homogenized by a motor-driven homogenizer (PowerGen125, Fisher Scientific, Chicago, IL) with 2-D lysis solution [7 M urea, 2 M thiourea, 4% (w/v) 3-[(3-cholamidopropyl)dimethyammonio]-1-propanesulfonate (CHAPS), 1% (w/v) dithiothreitol (DTT), 2% (v/v) Pharmalyte, and 1 mM benzamidine]. Proteins were extracted by vortexing for 30 min at room temperature and centrifugation at 15 000g for 30 min at 15 °C. Insoluble material was discarded and the soluble fraction was used for 2-D gel electrophoresis. For protein profiling of insoluble tissue fractions containing membrane or membrane-associated proteins, frozen tissue was homogenized in aqueous buffer solution (50 mM Tris-HCl, pH 7.5, 150 mM NaCl, and 1 mM benzamidin). After centrifugation at 12 000g for 1 h, the supernatant was discarded and the cell debris was used for extraction of membrane proteins. Cell debris was dissolved with 2-D lysis solution and proteins were extracted as described above. Protein concentrations of the tissue extracts were determined by Bradford assay.19 1158

Journal of Proteome Research • Vol. 9, No. 2, 2010

Kang et al. 2-D Gel Electrophoresis and Image Analysis. 2-D gel electrophoresis was performed as previously described.20 Briefly, protein extract was separated by isoelectric focusing (IEF) using an immobilized pH gradient (IPG) strip of nonlinear pH gradient of 4-10 (Genomine, Inc., Kyungbuk, Korea). Proteins were further separated by SDS-PAGE (26 × 20 cm format) for the second dimension. Proteins were detected by alkaline silver staining. Image analysis and quantification of protein spots was carried out using PDQuest software (BioRad, Hercules, CA). The quantity of protein in each spot was normalized by total valid spot intensity. Protein Profile Analysis. Clustering of samples and protein expression profiles was performed using Cluster and Tree View software (http://rana.Stanford.EDU/software/) to analyze and display differential patterns of expression profiles. Statistical significance of change in protein expression acquired from the average spot quantity between IZ and corresponding normal and tumor zones was evaluated by Student’s t test. Spots to be investigated were filtered by the significance of change of p < 0.05 and an expression ratio of at least (1.5. We applied the median centering for data adjustment and the default options of hierarchical clustering using the uncentered correlation similarity matrix. The distance between overall expression profiles for IZ with normal and tumor zones was determined as previously described,21 by summing the absolute values of the spot intensity ratios of proteins as given in the formula (a) where n is the number of included proteins and a and b are the normal, interface, or tumor samples.

∑ |log n

i)1

2

xjia xjib

|

((a))

The resulting distance matrix was used to build neighbor joining trees, as implemented in the MEGA software package (http://www.megasoftware.net). Protein Identification by MS. For protein identification by peptide mass fingerprinting, protein spots were excised, digested with trypsin (Promega, Madison, WI), mixed with R-cyano-4-hydroxycinnamic acid in 50% acetonitrile/0.1% TFA, and then subjected to MALDI-TOF analysis (Ettan MALDI-TOF, Amersham Biosciences, Piscataway, NJ), as previously described.22 For protein identification, ProFound (http://129.85.19.192/ profound_bin/WebProFound.exe) was used to search the NCBInr database. The following parameters were used for the database search: trypsin as cleaving enzyme, a maximum of 1 missed cleavage, iodoacetamide (Cys) as a complete modification, oxidation (Met) as a partial modification, monoisotopic masses, and a mass tolerance of (0.1 Da. Amino acid sequence-based protein identification by MALDI-PSD spectra of selected Nterminal derivatized peptides was carried out as previously described.23 Derivatization reactions were also performed as previously described.24 For protein identification, fragment masses obtained from MALDI-PSD were searched using Ettan MALDI-ToF software (Sonar) and/or the protein identification search engine PepFrag (http://prowl.rockefeller.edu/prowl/ pepfrag.html). Search inputs included measured precursors and fragment ion masses. The typical conservative error tolerance was (1.0 Da for the chemically averaged fragment masses. Immunofluorescent Staining, Confocal Microscopy, and Image Analysis. For immunofluorescent staining, we used same tissue samples as for MALDI-IMS and 2-D gel electro-

Molecular Proteomics Imaging of Tumor Interfaces

technical notes

Figure 1. MALDI ion images of differentially expressed proteins in ovarian tissues. Using the histological areas as defined by H&E staining, we imaged the tissue MS with PCA. (A) Optical H&E stained images. (B) All images of the differentially expressed proteins in ovarian cancer by MALDI-IMS. N, normal; T, tumor.

phoresis. Frozen tissue sections were fixed with 100% methanol. Sections were blocked in 0.1 mol/L phosphate buffer containing 3% bovine serum albumin and 0.1% Triton X-100, and incubated with both monoclonal anti-PRDX 1 (1:50 dilution, Abcam, Cambridge, MA) and polyclonal anti-plastin 2 (1:50 dilution, Abcam) antibodies in phosphate-buffered saline (PBS). The sections were incubated overnight at 4 °C. Sections were incubated in the dark for 1 h at room temperature with Alexa Fluor 594-conjugated goat anti-mouse IgG and Alexa Fluor 488conjugated goat anti-rabbit IgG (1:300 and 1:500, respectively, Molecular Probes, Leiden, Netherlands) secondary antibodies in blocking solution. The sections were mounted using ProLong Gold antifade reagent with DAPI (Molecular Probes).25 Antibody labeling was examined using a Zeiss LSM-510 laser scanning confocal microscope (Carl Zeiss, Thornwood, NY). The captured images were pseudocolored: red for rhodamine and green for FITC. Image analysis was performed using the standard system operating software provided with the Zeiss LSM-510 series microscope.

Results MALDI-IMS Analysis. We analyzed three ovarian cancer samples using MALDI-IMS. Tissue sections were coated with matrix and stained with H&E. Laser spot positions for each of the three samples (4329, 13013, and 3396 positions) covered the entire tissue sample, including tumor, normal and interface zones. Analysis of protein profiles to be displayed as high peaks ranged in spectral features from 3000 to 17 000 Da (Supple-

mentary Data Figure S1). To evaluate the quality of tissue spatial features based on MALDI-IMS data, we applied PCA statistical analysis (Supplementary Data Figure S2, Table S1). By combining direct tissue MS with PCA, we were able to extract signals with different localizations, and eventually images from each signal (Figure 1). Ions at 14 079, 4978, and 14 185 m/z were most abundant in interface regions, whereas those at 5001, 15 182, and 5034 m/z were most abundant in normal regions, and those at 15 262, 4978, and 11 464 m/z were found specifically within the tumor burden (Figure 2A,B). The boundary of the IZ between the tumor and normal zones was represented as merged. The merged images were displayed as red in the IZ, pink in the tumor, and green in the normal zone (Figure 2C). 2-D Gel Electrophoresis, Image Analysis, and Protein Identification by MS. The 2-DE patterns were performed to assess the molecular similarities or differences between the interface, normal, and tumor zones. A total of 106 differential protein spots showing expression changes between the interface, normal, and tumor zones were detected and represented for the histogram (Figure 3). Fifty spots showing quantitative protein expression changes were significantly selected more than 1.5-fold among the three sample groups (P < 0.05). A distance map tree was constructed to assess the similarities in protein expression between each group by adding the calculated distance of 50 spots as pairwise comparisons. The distances to IZs were closer to normal group whereas those to tumor samples were far from other groups (Figure 4A). IZs and Journal of Proteome Research • Vol. 9, No. 2, 2010 1159

technical notes

Kang et al.

Figure 2. MALDI-IMS images of differentially expressed proteins in ovarian tissues. (A) Optical H&E stained images. (B) Ions at 14 079, 13 994, and 11 464 m/z were abundant, particularly in the tumor zone. Ions at 5001, 15 182, and 5034 m/z were the most abundant in the normal zone and those at 15 262, 4978, and 14 185 m/z were the most abundant in IZ. (C) MALDI-IMS merged images of ovarian tissues shown in pseudocolor representation. Colors reflect intensities of selected mass signals. Red, interface zone; pink, tumor zone; green, normal zone. N, normal; T, tumor; IZ, interface zone.

Figure 3. 2-D gel images of ovarian tissue and nomination of spots using 2-D electrophoresis. (A-C) 2-D gel images of ovarian tissue and the normal, interface, and tumor zones. Each zone was categorized for protein expression comparisons between normal, IZ, and tumor zones. (D) A total of 106 differential protein spots were selected. The x-axis represents the pI of the protein and the y-axis represents the molecular weight of the protein. N, normal; T, tumor; IZ, interface zone.

normal samples demonstrated similarities in their protein expression profiles. Clustergram for previous profile distance 1160


analysis was constructed to assess the similarities in each protein expression. Three protein spots (#3005, #9026, #1619)

technical notes


Figure 4. Map tree, clustergram and identification of ovarian tissues expression profiles. (A) (a) Distance map tree representing the relative extents of expression changes between the tumor, normal, and interface zones. (b) Clustering of distance map tree. Each sample is categorized according to protein expression extents. (B) Clustergram was constructed from 50 differentially expressed spots between the tumor, normal, and interface zones. Green represents down-regulation, and red represents up-regulation. (C) Using the criteria of minimal 1.5-fold changes, differentially expressed protein spots from the 2-D gels were selected. (D) Peptide mass fingerprinting of three spots showing significant differences in 2-D gels. N, normal; T, tumor; IZ, interface zone. Table 1. Protein Identification of Interface Overexpressiona spot no.

protein name

accession no.

pIb/ pIc

MWb/ MWc (kDa)

sequence coverage (%)

est. Z score

fold IZ/Nd

fold IZ/Te

fold T/Nf

9026 1619

PRDX 1 Plastin 2

CAI13096 P13796

9.58/6.4 4.81/5.2

25.82/19.13 67.26/70.84

46 31

2.36 2.41

3.1 1.76

1.8 1.72

1.72 1.02

a The peptide profiles of the protein spots treated with trypsin were analyzed by MALDI-TOF MS. ProFound (http://129.85.19.192/profound_bin/ WebProFound.exe) was used to search the protein database for protein identification using peptide mass fingerprinting (PMF). The mass and pI values specified are theoretically matched by a database search. b Observed. c Theoretically calculated. d The rate of increase in intensity (average interface sample intensity/average normal sample intensity). e The rate of increase in intensity (average interface sample intensity/average tumor sample intensity). f The rate of increase in intensity (average tumor sample intensity/average normal sample intensity).

in IZs were found to be expressed at higher levels than in normal and tumor zones (Figure 4B). On peptide mass fingerprinting using MALDI-TOF, 2 of the 3 protein spots were finally identified (Figure 4C,D and Table 1). MALDI-TOF MS analysis indicated that those specific to IZs were identified as plastin 2 and PRDX 1. Immunofluorescent Confocal Microscopy. Immunofluorescent staining revealed a significant overexpression of plastin 2 and PRDX 1 within IZ of ovarian tissues. IZ was sharply

delineated in plastin 2 overexpression. Otherwise, PRDX 1 demonstrated gradual increase of fluorescent signals in IZ (Figure 5).

Discussion Recent reports indicate that MALDI-IMS is an innovative tool that can be used for protein identification in crude tissue samples.8,26,27 Current protein identification methods are based on extraction, proteolysis, and liquid chromatography (LC)/ Journal of Proteome Research • Vol. 9, No. 2, 2010 1161

technical notes

Kang et al.

Figure 5. Plastin 2 and PRDX 1 immunofluorescence in ovarian tissues. (A) Plastin 2 and (B) PRDX 1 expression in ovarian cancer tissue by immunofluorescent confocal microscopy. Notice restricted fluorescent signals of both plastin 2 and PRDX 1 along IZ. N, normal; T, tumor; IZ, interface zone. Scale bar ) 100 µm.

MS/MS techniques. However, the MALDI-IMS protocol used in this study allows for protein identification directly from tissues, while preserving the spatial integrity of the tissue sample. The MALDI-IMS procedure using tissue samples is diagramed in Figure 6, and can be summarized by the following steps: (1) Application of matrix and enzymatic digestion of proteins; (2) identification of proteins using protein database matching; (3) validation of protein identities. Several conditions need to be optimized for successful tissue MALDI-IMS, including enzymatic digestions, solvent composition, and matrix composition. For MALDI-IMS analysis of peptides and proteins, as for standard MALDIIMS, the choice of matrix is generally determined by the mass size range. CHCA (R-Cyano-4-hydroxycinnamic acid) is the matrix of choice for detection of low molecular weight peptides using a TOF mass spectrometer in reflectron mode. SA (3,5-dimethoxy-4-hydroxycinnamic acid) is preferentially used for relatively higher masses using the linear mode. Crystallization characteristics of the common matrices used on tissue sections was reviewed early in the development of MALDI-IMS protocols.14,15,28 Protein analysis is most frequently carried out with tissues prepared using SA as the matrix in 50-60% acetonitrile.29 This protocol tends to give the best protein extraction, sensitivity, and resolution for high molecular mass species (>5000 Da). Image analysis of discrete molecules in tissue can be acquired by using MALDI-IMS to determine their spatial localization with a lateral resolution of 80-200 µm. To evaluate the quality of tissue spatial features based on MALDI-IMS data, we applied PCA statistical analysis. PCA provided information about which molecules were altered in the tissue. By combining direct tissue MS with PCA, we were able to extract signals with different localizations and eventually images for each signal.17,18 MALDI-IMS has been used to study the molecular aspects of cancer, providing tumor-specific, as well as diagnostic markers.30,31 The identification of biomarkers directly from tissue sections is important for the diagnosis of tumors, and is the main aim for developing the MALDI-IMS technology. Additionally, MALDI-IMS may have an even greater potential for providing more complete tissue profiles.32 1162


Pathological signatures of EMT or invasion markers in tumor tissues are most dynamic and active within the tumor front directly invading the normal microenvironment.2 Each tumor microenvironment is unique, and the zone between the tumor and normal stroma is not only a geographical region, but also a unique functional and molecular region. The IZ, the region between the tumor invasion front and the normal zone, develops during tumor invasion and metastasis through remodeling of the tumor microenvironment (TME). Considering that most tumor burdens follow the activation of TME remodeling, normal zone that is remote from the tumor load may yet be unaffected. Theoretically, remodeling of the TME could occur most rapidly in the IZ. Large-scale cDNA array and tissue microarray-based immunohistochemistry studies support the existence of an interface that could be directly associated with tumor invasion and metastasis.2,33,34 In this study, we used MALDI-IMS to identify ovarian tumor profiles, and specifically to identify interface biomarkers. Using MALDI-IMS technology, we confirmed that IZs have unique molecular profiles. By classifying the three zones, central tumor burden, remote normal zone, and IZs, based on pathological findings, we confirmed that IZs are different from the tumor or normal zones at the molecular level. In addition, immunofluorescence targeting of plastin 2 and PRDX 1 in ovarian tissue sections also confirmed their localization along the IZ. We found that specifically plastin 2 and PRDX 1 are upregulated in IZs. Plastin 2 gene expression have been reported to be significantly correlated with the progression of cancer staging and a potential metastatic marker.35 PRDX1 is suspected to be associated with malignant and transformation in cancer.36 This protein may have a proliferative effect and play a role in cancer development or progression. PRDX1 encodes a member of the peroxiredoxin family of antioxidant enzymes and may play an antioxidant protective role in cells. In conclusion, IZs are different from the adjacent tumor or normal zones, and plastin 2 and PRDX 1 may be interfacespecific markers for comparative proteomics. Abbreviations: MALDI, matrix-assisted laser desorption/ ionization; MS, mass spectrometry; IMS, MALDI-imaging MS;

technical notes


References

Figure 6. Schematic representation of the MALDI-IMS technology. The protocol of MALDI-IMS using tissue samples can be summarized as follows: (1) Application of matrix and enzymatic digestion of proteins. (2) Identification of proteins using protein database matching. (3) Validation of protein identities. The protein peaks of MALDI-IMS are detected form 5 to 30 kDa protein. To detect larger proteins (>30 kDa) by MALDI-IMS, the larger proteins may be enabled to digest by the in situ enzymatic digestion on the tissue. In situ enzymatic digestion on the tissue is needed to determine the optimal enzyme conditions that vary enzyme concentration, time, volume, and solvent composition. For MALDI-MSI of peptides and proteins, the choice of matrix is normally determined by the mass range. For lower molecular weight masses (peptides), typically CHCA and DHB matrix are used. For higher molecular masses (proteins), SA matrix is preferred. There are two ways to identify protein. One way is 2-D MALDI-TOF MS (represented in a solid line) and another is nanoliqid chromatography (LC) MS (shown in dotted line). Recently, identification of proteins may be directly performed from the MALDI-IMS without any analytic process of extraction and proteolysis.16,31 CHCA, R-Cyano-4-hydroxycinnamic acid; DHB, 2,5-dihydroxybenzoic acid; SA, sinapinic acid; MS, mass spectrometry; solid line, performed in the present study; dotted line, not performed in the present study, but possible.

PCA, principal component analysis; EMT, epithelial-mesenchymal transition; IF, interface; PRDX 1, peroxiredoxin 1.

Acknowledgment. This study was supported by the Korea Healthcare Technology R&D Project, Ministry for Health, Welfare and Family Affairs, Republic of Korea (CNH A084550) and by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MEST) (No. 20090079165; CNH). Supporting Information Available: Figure S1, laser spot positions and mass spectrum image processing. Figure S2, PCA of ovarian cancer data set. Table S1, mass list of PCA. This material is available free of charge via the Internet at http://pubs.acs.org.

(1) Shih, I.-M.; Kurman, R. J. Ovarian tumorigenesissa proposed model based on morphological and molecular genetic analysis. Am. J. Pathol. 2004, 164, 1511–1518. (2) Wever, O. D.; Pauwels, P.; Craene, B. D.; Sabbah, M.; Emami, S.; Redeuilh, G.; Gespach, C.; Bracke, M.; Berx, G. Molecular and pathological signatures of epithelial-mesenchymal transitions at the cancer invasion front. Histochem. Cell Biol. 2008, 130, 481– 494. (3) Lahm, H. W.; Langen, H. Mass spectrometry: a tool for the identification of proteins separated by gels. Electrophoresis 2000, 21, 2105–2114. (4) Pandey, A.; Mann, M. Proteomics to study genes and genomes. Nature 2000, 405, 837–846. (5) Aebersold, R.; Goodlett, D. R. Mass spectrometry in proteomics. Chem. Rev. 2001, 101, 269–295. (6) Caprioli, R. M.; Farmer, T. B.; Gile, J. Molecular imaging of biological samples: Localization of peptides and proteins using MALDI-TOF MS. Anal. Chem. 1997, 69, 4751–4760. (7) Chaurand, P.; Stoeckli, M.; Caprioli, R. M. Direct profiling of proteins in biological tissue sections by MALDI mass spectrometry. Anal. Chem. 1999, 71, 5263–5270. (8) Stoeckli, M.; Chaurand, P.; Hallahan, D. E.; Caprioli, R. M. Imaging mass spectrometry: A new technology for the analysis of protein expression in mammalian tissues. Nat. Med. 2001, 7, 493–496. (9) Chaurand, P; Schwartz, S. A.; Caprioli, R. M. Imaging mass spectrometry: A new tool to investigate the spatial organization of peptides and proteins in mammalian tissue sections. Curr. Opin. Chem. Biol. 2002, 6, 676–681. (10) Chaurand, P.; Caprioli, R. M. Direct profiling and imaging of peptides and proteins from mammalian cells and tissue sections by mass spectrometry. Electrophoresis 2002, 23, 3125–3135. (11) Khatib-Shahidi, S.; Andersson, M.; Herman, J. L.; Gillespie, T. A.; Caprioli, R. M. Direct molecular analysis of whole-body animal tissue sections by imaging MALDI mass spectrometry. Anal. Chem. 2006, 78, 6448–6456. (12) Reyzer, M. L.; Caprioli, R. M. MALDI-MS-based imaging of small molecules and proteins in tissues. Curr. Opin. Chem. Biol. 2007, 11, 29–35. (13) Shimma, S.; Furuta, M.; Ichimura, K.; Yoshida, Y.; Setou, M. A novel approach to in situ proteome analysis using a chemical inkjet printing technology and MALDI-QIT-TOF tandem mass spectrometer. J. Mass Spectrom. Soc. Japan 2006, 54, 133–140. (14) Schwartz, S. A.; Reyzer, M. L.; Caprioli, R. M. Direct tissue analysis using matrix-assisted laser desorption/ ionization mass spectrometry: practical aspects of sample preparation. J. Mass Spectrom. 2003, 38, 699–708. (15) Crecelius, A. C.; Cornett, D. S.; Caprioli, R. M.; Williams, B.; Dawant, B. M.; Bodenheimer, B. Three-dimensional visualization of protein expression in mouse brain structures using imaging mass spectrometry. Am. Soc. Mass Spectrom. 2005, 16, 1093–1099. (16) Goodwin, R. J.; Pennington, S. R.; Pitt, A. R. Protein and peptides in pictures: Imaging with MALDI mass spectrometry. Proteomics 2008, 8, 3785–3800. (17) Gerhard M.; Deininger S. O.; Schleif F. M. Statistical Classification and Visualization of MALDI-Imaging Data. In Proceedings of the Twentieth IEEE International Symposium on Computer-Based Medical Systems; Los Alamitos, 2007; pp 403-405. (18) Yao, I.; Sugiura, Y.; Matsumoto, M.; Setou, M. In situ proteomics with imaging mass spectrometry and principal component analysis in the Scrapper-knockout mouse brain. Proteomics 2008, 8, 3692– 3701. (19) Bradford, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 1976, 72, 248–254. (20) Rabilloud, T.; Kieffer, S.; Procaccio, V; Louwagie, M.; Courchesne, P. L.; Patterson, S. D.; Martinez, P.; Garin, J.; Lunardi, J. Twodimensional electrophoresis of human placental mitochondria and protein identification by mass spectrometry: toward a human mitochondrial proteome. Electrophoresis 1998, 9, 1006–1014. (21) Enard, W.; Khaitovich, P.; Klose, J.; Zo¨llner, S.; Heissig, F.; Giavalisco, P.; Nieselt-Struwe, K.; Muchmore, E.; Varki, A.; Ravid, R.; Doxiadis, G. M.; Bontrop, R. E.; Pa¨a¨bo, S. Intra- and interspecific variation in primate gene expression patterns. Science 2002, 296, 340–343. (22) Fernandez, J.; Gharahdaghi, F.; Mische, S. M. Routine identification of proteins from sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels or polyvinyl difluoride membranes using matrix assisted laser desorption/ionization-time of flightmass spectrometry (MALDI-TOF-MS). Electrophoresis 1998, 19, 1036–1045.

Journal of Proteome Research • Vol. 9, No. 2, 2010 1163

technical notes (23) Flensburg, J.; Belew, M. Characterization of recombinant human serum albumin using matrix-assisted laser desorption ionization time-of-flight mass spectrometry. J. Chromatogr., A 2003, 1009, 111–117. (24) Wang, D.; Kalb, S. R.; Cotter, R. J. Improved procedures for N-terminal sulfonation of peptides for matrix-assisted laser desorption/ionization post-source decay peptide sequencing. Rapid Commun. Mass Spectrom. 2004, 18, 96–102. (25) Goedert, M.; Spillantini, M. G.; Cairns, N. J.; Crowther, R. A. Tau proteins of Alzheimer paired helical filaments: abnormal phosphorylation of all six brain isoforms. Neuron 1992, 8, 159–168. (26) Herring, K. D.; Oppenheimer, S. R.; Caprioli, R. M. Direct tissue analysis by MALDI MS: application to kidney biology. Semin. Nephrol. 2007, 27, 597–608. (27) Chaurand, P.; Norris, J. L.; Cornett, D. S.; Mobley, J. A.; Caprioli, R. M. New developments in profiling and imaging of proteins from tissue sections by MALDI mass spectrometry. J. Proteome Res. 2006, 5, 2889–2900. (28) Schwartz, S. A.; Weil, R. J.; Johnson, M. D.; Toms, S. A.; Caprioli, R. M. Protein profiling in brain tumors using mass spectrometry: feasibility of a new technique for the analysis of protein expression. Clin. Cancer Res. 2004, 10, 981–987. (29) Schwartz, S. A.; Weil, R. J.; Thompson, R. C.; Shyr, Y.; Moore, J. H.; Toms, S. A.; Johnson, M. D.; Caprioli, R. M. Proteomic-based prognosis of brain tumor patients using direct-tissue matrixassisted laser desorption ionization. Cancer Res. 2005, 65, 7674– 7681. (30) Chaurand, P.; Schriver, K. E.; Caprioli, R. M. Instrument design and characterization for high resolution MALDI-MS imaging of tissue sections. J. Mass Spectrom. 2007, 42, 476–489.

1164


Kang et al. (31) Seeley, E. H.; Caprioli, R. M. Molecular imaging of proteins in tissues by mass spectrometry. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 8126–18131. (32) Maddalo, G.; Petrucci, F.; Iezzi, E; Pannellini, T. Analytical assessment of MALDI-TOF imaging mass spectrometry on thin histological samples. An insight in proteome investigation. Clin. Chim. Acta 2005, 357, 210–218. (33) Alonso, S. R.; Tracey, L.; Ortiz, P.; Perez-Gomez, B.; Palacios, J.; Pollan, M.; Linares, J.; Serrano, S.; Saez-Castillo, A. I.; Sanchez, L.; Pajares, R.; Sanchez-Aguileran, A.; Artiga, M. J.; Piris, M. A.; Rodriguez-Peralto, J. L. A high-thoughput study in melanoma identifies epithelial-mesenchymal trnasition as a mojor determinants of metastasis. Cancer Res. 2007, 67, 3450–3460. (34) Vasco, V.; Espinosa, A. V.; Scouten, W.; He, H.; Auer, H.; Liyanarachchi; Larin, A.; Savchenki, V; Francis, G. L.; de la Chapelle, A.; Saji, M; Ringel, M. D. Gene expression and functional evidence of epithelial-to mesenchymal transition in papillary thyroid carcinoma invasion. Proc. Natl. Acad. Sci. U.S.A. 2007, 104, 2803– 2808. (35) Otsuka, M.; Kato, M.; Yoshikawa, T.; Chen, H.; Brown, E. J.; Masuho, Y.; Omata, M.; Seki, N. Differential expression of the L-plastin gene in human colorectal cancer progression and metastasis. Biochem. Biophys. Res. Commun. 2001, 289, 876–881. (36) Demasi, A. P.; Furuse, C.; Soares, A. B.; Altemani, A.; Arau ´ jo, V. C. Peroxiredoxin I, platelet-derived growth factor A, and plateletderived growth factor receptor R are overexpressed in carcinoma ex pleomorphic adenoma: association with malignant transformation. Hum. Pathol. 2009, 40, 390–397.

PR900666Q

Molecular Proteomics Imaging of Tumor Interfaces by Mass

Recommend Documents