Table of Contents
Related Titles
Title page
Copyright page
Preface to the Second Edition
List of Contributors
1: The MALDI Process and Method
1.1 Introduction
1.2 Analyte Incorporation
1.3 Absorption of the Laser Radiation
1.4 The Ablation/Desorption Process
1.5 Ionization
1.6 Fragmentation of MALDI Ions
1.7 MALDI of Noncovalent Complexes
1.8 The Optimal Choice of Matrix: Sample Preparation
Abbreviations
2: MALDI Mass Spectrometry Instrumentation
2.1 Introduction
2.2 Lasers for MALDI-MS
2.3 Fragmentation of MALDI Ions
2.4 Mass Analyzers
2.5 Fourier Transform Ion Cyclotron Resonance Mass Spectrometers
2.6 Quadrupole Ion Trap Mass Spectrometers
2.7 Hybrid Mass Spectrometers
2.8 Future Directions
Definitions and Acronyms
3: MALDI-MS in Protein Chemistry and Proteomics
3.1 Introduction
3.2 Sample Preparation for Protein and Peptide Analysis by MALDI-MS
3.3 Strategies for Using MALDI-MS in Protein Biochemistry
3.4 Applications of MALDI-MS in Proteomics
3.5 Computational Tools for Protein Analysis by MALDI-MS
3.6 Clinical Applications of MALDI-MS
3.7 Conclusions
Acknowledgments
4: MALDI-Mass Spectrometry Imaging
4.1 Introduction
4.2 History of Mass Spectrometry Imaging (MSI) and Microprobing Techniques
4.3 MALDI in Micro Dimensions: Instruments and Mechanistic Differences
4.4 Visualization of Mass Spectrometric Information
4.5 Data Processing and Data Exchange
4.6 Matrix Deposition for High-Resolution Imaging
4.7 Organisms, Organs, and Tissues: MALDI Imaging at Various Lateral Resolutions
4.8 Whole-Cell and Single-Cell Analysis
4.9 Cell Sorting and Capturing
4.10 Direct Protein Identification and Localization
4.11 Identification and Characterization: Requirements for Mass Resolution and Accuracy
4.12 Conclusions
Acknowledgments
5: Analysis of Nucleic Acids, and Practical Implementations in Genomics and Genetics
5.1 Challenges in Nucleic Acid Analysis by MALDI-MS
5.2 Genetic Markers
5.3 Assay Formats for Nucleic Acid Analysis by MALDI-MS
5.4 Applications in Genotyping
5.5 Applications in Comparative Sequence Analysis
5.6 Applications in Quantitation of Nucleic Acids for Analysis of Gene Expression and Gene Amplification
5.7 Future Perspectives for the MALDI-MS Analysis of Nucleic Acids
Acknowledgments
6: MALDI-MS of Glycans and Glycoconjugates
6.1 Introduction
6.2 Profiling of Glycans and Glycosphingolipids
6.3 Structural Determination
6.4 Quantitative Analysis
6.5 Conclusions
7: Lipids
7.1 Introduction
7.2 Analysis of Individual Lipid Classes and Their Characteristics
7.3 MALDI-TOF-MS of Typical Lipid Mixtures
7.4 Characterization of Typical Oxidation Products of Lipids
7.5 MALDI-MS Imaging
7.6 Combining TLC and MALDI for Lipid Analysis
7.7 Summary and Outlook
Acknowledgments
Abbreviations
8: MALDI-MS for Polymer Characterization
8.1 Introduction
8.2 Technical Aspects of MALDI-MS
8.3 Attributes and Limitations of MALDI-MS
8.4 Conclusions and Perspectives
9: Small-Molecule Desorption/Ionization Mass Analysis
9.1 Introduction
9.2 Matrix Choices for Small-Molecule MALDI
9.3 Sample Preparation
9.4 Qualitative Characterization of LMM Molecules
9.5 Analyte Quantitation by MALDI
9.6 Conclusions
Acknowledgments
Abbreviations/Acronyms
10: Computational Analysis of High-Throughput MALDI-TOF-MS-Based Peptide Profiling
10.1 Introduction
10.2 MALDI-MS Data Preprocessing
10.3 Statistical Analysis of Preprocessed Data
10.4 Concluding Remarks
11: Biotyping of Microorganisms
11.1 The Technique
11.2 Standard Identification of Bacteria and Other Microorganisms
11.3 Applicability and Performance in Routine Laboratories
11.4 Direct Specimen Analysis
11.5 Subtyping
11.6 Resistance Testing
11.7 Outlook
Index
Related Titles
Editors
Prof. Dr. Franz Hillenkamp
Institute for Medical Physics
University of Münster
Robert-Koch-Str. 31
48149 Münster
Germany
Prof. Dr. Jasna Peter-Katalinic
Department of Biotechnology
University of Rijeka
Radmile Matejčić 2
51000 Rijeka
Croatia
Cover
High speed time lapse photograph of IR-MALDI plumes generated with an optical parametric oscillator (OPO) laser (for more details see Fig. 1.2)
Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty can be created or extended by sales representatives or written sales materials. The Advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.
Library of Congress Card No.: applied for
British Library Cataloguing-in-Publication Data
A catalogue record for this book is available from the British Library.
Bibliographic information published by the Deutsche Nationalbibliothek
The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available on the Internet at <http://dnb.d-nb.de>.
© 2014 Wiley-VCH Verlag GmbH & Co. KGaA, Boschstr. 12, 69469 Weinheim, Germany
Wiley-Blackwell is an imprint of John Wiley & Sons, formed by the merger of Wiley's global Scientific, Technical, and Medical business with Blackwell Publishing.
All rights reserved (including those of translation into other languages). No part of this book may be reproduced in any form – by photoprinting, microfilm, or any other means – nor transmitted or translated into a machine language without written permission from the publishers. Registered names, trademarks, etc. used in this book, even when not specifically marked as such, are not to be considered unprotected by law.
Print ISBN: 978-3-527-33331-8
ePDF ISBN: 978-3-527-67373-5
ePub ISBN: 978-3-527-67374-2
Mobi ISBN: 978-3-527-67372-8
oBook ISBN: 978-3-527-33596-1
Typesetting Toppan Best-set Premedia Limited, Hong Kong
Cover Design Adam-Design, Weinheim, Germany
Preface to the Second Edition
This book on matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS), first published in 2006, has obviously fulfilled a long felt demand among the community of bioorganic mass spectrometrists. It was sold out after only a few years. To prepare a second edition has been a considerable task. MALDI-MS is still a very active and developing field, requiring essential changes and additions to the first edition while keeping it to a handy size and particularly staying truthful to the concept of it being a “practical guide” more than an in-depth treatment of the basics.
Chapter 1 has been amended by the results of Karas and Jaskolla on new design matrices and the mechanisms of ion formation in MALDI, which has substantially added to our understanding of the processes and how to optimize them for practical applications. They have also influenced our view of the method as such and have led to revision of other parts of the chapter.
While most of the instruments described in the first edition are still in use in many laboratories, two newcomers have literally revolutionized particularly the routine applications: the orbitrap and the ion mobility instruments. Both are now covered in much detail in Chapter 2.
Proteomics had already been at a rather mature state at the time the first edition was published. Most additions and improvements in this field are somewhat special to a given problem and are not covered in detail in this book. Here, as in most of the other applications, the reader is referred to the extensive list of original literature at the end of each chapter.
Chapter 4 has been essentially rewritten. It now concentrates on MALDI imaging, a field which has seen a dramatic development in recent years and promises to continue on this path. The discussion of biomarkers has been referred to the new Chapter 10 on bioinformatics. MALDI imaging requires the treatment of the raw data with rather sophisticated software tools, specific for this application. It is, therefore, contained in Chapter 4, rather than Chapter 10.
MALDI-MS of nucleic acids, covered in Chapter 5, has still not found as widespread an application as, for example, the analysis of proteins, mostly because of competing techniques such as second generation sequencers which are now in routine use. One interesting MALDI application is the analysis of RNA and other modified nucleic acids, where straight sequencing leads to a loss of important information.
Application of MALDI-MS to analysis of protein-linked N- and O-glycans in Chapter 6 has been generally revised and updated. In consideration of growing attention to glycomics in biology and medicine, efficient protocols for glycans and glycopeptides have been described to encompass the carbohydrate complexity both for rapid mapping as well as for quantification, and those for glycosphingolipids added.
Revisions in Chapter 7 are related to the increased interest in lipid analysis, notably boosted by introduction of the new “omics” field – lipidomics – and by developments of MS imaging as a robust new application of MALDI-MS. In addition, novel potentials of lipid analysis in applications of the direct desorption from solid surfaces and MALDI-MS imaging to diagnostics using lipids as disease markers are described.
Chapter 8, on the analysis of synthetic polymers, remained essentially unchanged in the second edition.
Chapter 9, on small molecule desorption/ionization mass analysis, reflects the use of MALDI-MS in the development of new pharmaceutical agents, again a field of important applications and developments.
Bioinformatics, now covered in the new Chapter 10, was obviously missing in the first edition. Thang V. Pham and Connie R. Jimenez of the Free University of Amsterdam describe how they use bioinformatics software in their search for tumor markers in human samples. While this is a rather special application, they have taken great care to refer the reader to the original literature which describes the principles in other fields of application.
Biotyping of microorganisms, covered in Chapter 11, has been added as a last minute topic. This chapter is not as comprehensive as the others, but we considered it important, because it is the first and so far only large scale routine clinical MALDI application. It has boomed over the last two years and is still in a phase of intense development. The most important aspects of this new application are discussed in the chapter; for details, readers are referred to the many listed references.
Franz Hillenkamp
Jasna Peter-Katalinic
Münster, Rijeka, April 2013
List of Contributors
Stefan Berkenkamp
SEQUENOM Inc.
3595 John Hopkins Court
San Diego, CA 92121
USA
sberkenkamp@sequenom
Dirk van den Boom
SEQUENOM Inc.
3595 John Hopkins Court
San Diego, CA 92121
USA
dvandenboom@sequenom.com
Lucinda H. Cohen
Merck Research Laboratories
DMPK Bioanalytical Group
Mail Stop RY800B201
Rahway, NJ 07065
USA
lucinda_cohen@merck.com
Klaus Dreisewerd
University of Münster
Institute for Medical Physics and Biophysics
Robert-Koch-Str. 31
48149 Münster
Germany
dreisew@uni-muenster.de
Daniele Fabris
University at Albany
The RNA Institute
1400 Washington Avenue
Albany, NY 1222
USA
dfabris@albany.edu
Beate Fuchs
University of Leipzig
Institute of Medical Physics and Biophysics, Faculty of Medicine
Härtelstr. 16–18
04109 Leipzig
Germany
beate.fuchs@medizin.uni-leipzig.de
Eden P. Go
Department of Chemistry
University of Kansas
Lawrence, KS 66047
USA
edenp@ku.edu
Franz Hillenkamp
University of Münster
Institute for Medical Physics and Biophysics
Robert-Koch-Str. 31
48149 Münster
Germany
hillenk@uni-muenster.de
Karin Hjernø
University of Southern Denmark
Department of Biochemistry and Molecular Biology
Campusvej 55
5230 Odense
Denmark
hjernoe@bmb.sdu.dk
Thorsten W. Jaskolla
University of Münster
Institute for Medical Physics and Biophysics
Robert-Koch-Str. 31
D-48149 Münster
Germany
tjaskolla@uni-muenster.de
Ole N. Jensen
University of Southern Denmark
Department of Biochemistry and Molecular Biology
Campusvej 55
5230 Odense
Denmark
jenseno@bmb.sdu.dk
Connie R. Jimenez
VU University Medical Center
VUmc-Cancer Center Amsterdam, Department of Medical Oncology
CCA 1-46, OncoProteomics Laboratory
De Boelelaan 1117
1081 HV Amsterdam
The Netherlands
c.jimenez@vumc.nl
Michael Karas
Johann Wolfgang Goethe University of Frankfurt
Institute of Pharmaceutical Chemistry, Biocenter
Max-von-Laue-Str. 9
60438 Frankfurt am Main
Germany
karas@pharmchem.uni-frankfurt.de
Markus Kostrzewa
Vice President – Clinical Mass Spectrometry
Bruker Daltonik GmbH
Fahrenheitstr. 4
28359 Bremen
Germany
marcus.kostrzewa@bdal.de
Erika Lattová
Department of Chemistry
University of Manitoba
144 Dysart Road
Winnipeg, MB R3T 2N2
Canada
lattovae@cc.umanitoba.ca
Fangbiao Li
Merck Research Laboratories
DMPK Bioanalytical Group
Mail Stop RY800B201
Rahway, NJ 07065
USA
fangbiao.li@merck.com
Liang Li
University of Alberta
Department of Chemistry
Chemistry Centre W3-39
Edmonton, AB T6G 2G2
Canada
liang.li@ualberta.ca
Peter B. O'Connor
University of Warwick
Department of Chemistry
Gibbet Hill Road
Coventry CV4 7AL
UK
p.oconnor@warwick.ac.uk
Hélène Perreault
Department of Chemistry
University of Manitoba
144 Dysart Road
Winnipeg, MB R3T 2N2
Canada
helene.perreault@umanitoba.ca
Jasna Peter-Katalinic
Department of Biotechnology
University of Rijeka
Radmile Matejčić 2
51000 Rijeka
Croatia
jasnapk@biotech.uniri.hr
Thang V. Pham
VU University Medical Center
VUmc-Cancer Center Amsterdam, Department of Medical Oncology
CCA 1-46, OncoProteomics Laboratory
De Boelelaan 1117
1081 HV Amsterdam
The Netherlands
t.pham@vumc.nl
Dijana Šagi
Sanofi-Aventis Deutchland GmbH
Industriepark Höchst, Geb. H773
65926 Frankfurt am Main
Germany
Dijana.Sagi@sanofi.com
Jürgen Schiller
University of Leipzig
Institute of Medical Physics and Biophysics, Faculty of Medicine
Härtelstr. 16–18
04109 Leipzig
Germany
juergen.schiller@medizin.uni-leipzig.de
Gary Siuzdak
The Scripps Research Institute
Center for Metabolomics and Mass Spectrometry
Departments of Chemistry, Molecular and Computational Biology
BCC007
10550 North Torrey Pines Road
La Jolla, CA 92037
USA
siuzdak@scripps.edu
Bernhard Spengler
Justus Liebig University Giessen
Institute of Inorganic and Analytical Chemistry
Schubertstr. 60, Bldg 16
35392 Giessen
Germany
Bernhard.Spengler@anorg.chemie.uni-giessen.de
Kerstin Strupat
Thermo Fisher Scientific
Life Science Mass Spectrometry
Hanna-Kunath-Str. 11
28199 Bremen
Germany
kerstin.strupat@thermofisher.com
1
The MALDI Process and Method
Matrix-assisted laser desorption/ionization (MALDI) is one of the two “soft” ionization techniques besides electrospray ionization (ESI) which allow for the sensitive detection of large, nonvolatile and labile molecules by mass spectrometry. Over the past 27 years, MALDI has developed into an indispensable tool in analytical chemistry, and in analytical biochemistry in particular. In this chapter, the reader will be introduced to the technology as it stands now, and some of the underlying physical and chemical mechanisms as far as they have been investigated and clarified to date will be discussed.
Attention will also be focused on the central issues of MALDI, that are necessary for the user to understand for the efficient application of this technique. As an in-depth discussion of these topics is beyond the scope of this chapter, the reader is referred to recent reviews [1–4]. Details of the current state of instrumentation, including lasers and their coupling to mass spectrometers, will be presented in Chapter 2.
As with most new technologies, MALDI came as rather a surprise even to the experts in the field on the one hand, but also evolved from a diversity of prior art and knowledge on the other hand. The original notion had been that (bio)molecules with masses in excess of about 500–1000 Da could not be isolated out of their natural (e.g., aqueous) environment, and even less be charged for an analysis in the vacuum of a mass spectrometer without excessive and unspecific fragmentation. During the late 1960s, however, Beckey introduced field desorption (FD), the first technique to open a small road into the territory of mass spectrometry (MS) of bioorganic molecules [5]. Next came secondary ion mass spectrometry (SIMS), and in particular static SIMS, as introduced by A. Benninghoven in 1975 [6]. This development was taken a step further by M. Barber in 1981, with the bombardment of organic compounds dissolved in glycerol with high-energy atoms, which Barber coined fast atom bombardment (FAB). It was in this context, and in conjunction with the first attempts to desorb organic molecules with laser irradiation, that the concept of a “matrix” as a means of facilitating desorption and enhancing ion yield was born [7]. The principle of desorption by the bombardment of organic samples with the fission products of the 252Cf nuclear decay, later termed plasma desorption (PD), was first described by R. Macfarlane in 1974 [8]. Subsequently, the groups of Sundqvist and Roepstoff greatly improved the analytical potential of this technique by the addition of nitrocellulose, which not only cleaned up the sample but was also suspected of functioning as a signal-enhancing matrix [9].
The first attempts at using laser radiation to generate ions for a mass spectrometric analysis were reported only a few years after the invention of the laser [10, 11]. Vastola and Pirone had already demonstrated the possibility of recording the spectra of organic compounds with a time-of-flight (TOF) mass spectrometer. Subsequently, several groups continued to pursue this line of research, mainly R. Cotter at Johns Hopkins University in the USA and P. Kistemaker at the FOM Institute in Amsterdam, the Netherlands. Indeed, for a number of years the Amsterdam group held the high-mass record for a bioorganic analyte with a spectrum of underivatized digitonin at mass 1251 Da ([M + Na]+), desorbed with a CO2-laser at a wavelength of 10.6 μm in the far infrared (IR) [12].
Independently of, and parallel to, these groups, Hillenkamp and Kaufmann developed the laser microprobe mass analyzer (LAMMA) [13], the commercial version of which was marketed by Leybold Heraeus in Cologne, Germany and which is now on exhibition in the section on New Technologies of the Deutsches Museum in Munich, Germany. The instrument originally comprised a frequency-doubled ruby laser at a wavelength of 347 nm in the near ultraviolet (UV), and later a frequency-quadrupled Nd:YAG-laser at a wavelength of 266 nm in the far UV. The laser beam was focused to a spot of ≤1 μm in diameter to probe thin tissue sections for inorganic ions and trace elements such as Na, K, and Fe. The mass analyzer of the LAMMA instruments was also a TOF mass spectrometer, and was the first commercial instrument with an ion reflector, which had been invented a few years earlier by B.A. Mamyrin in Leningrad. The sensitivity-limiting “noise” of the LAMMA spectra were signals that were soon identified as coming from the organic polymer used to embed the tissue sections, as well as other organic tissue constituents. It was this background noise which triggered the search for a systematic analysis of organic samples and which eventually led to the discovery of the MALDI principle in 1984. The principle and its acronym were first described in 1985 [14], and the first spectrum of the nonvolatile bee venom mellitin, an oligopeptide at mass 2845 Da, in 1986 [15]. Spectra of proteins with masses exceeding 10 kDa and 100 kDa were reported in 1988 [16], and details presented at the International Mass Spectrometry Conference in Bordeaux in 1988, respectively.
Both, ESI and MALDI were developed independently but concurrently, and when their potential for the desorption of nonvolatile, fragile (bio)molecules was discovered, the scientific community was mostly impressed by the ability of these techniques to access the high mass range, particularly of proteins. However, FAB- and PD-MS had at that time already generated spectra of trypsin at mass 23 kDa and other high-mass proteins. What really made the difference in particular for the biologists was the stunning sensitivity which, for the first time, made MS compatible with sample preparation techniques used in these fields. For MALDI, the minimum amount of protein needed for a spectrum of high quality was reduced from 1 pmol in 1988 to a few femtomoles only about a year later. Today, in favorable cases, the level is now down in the low attomole range. Many other developments – both instrumental (see Chapter 2) as well as specific sample preparation recipes and assays (see other chapters of the book) – took place during the following decade, and the joint impact of all of these together has today made MALDI-MS an indispensable tool not only in the life sciences but also in polymer analysis, food sciences, pharmaceutical drug discovery, or forensic jurisprudence.
The use of a chemical matrix in the form of small, laser-absorbing organic molecules in large excess over the analyte is at the core of the MALDI principle. Several developments for laser desorption schemes took place in parallel to and following publications of the MALDI principle. These all attempted to replace the chemical matrix by a more easy-to-handle physical matrix, or a more simple combination of the two. The best known of these was the system of Tanaka and coworkers, which was first presented at a Sino-Japanese conference in 1987; the details were subsequently published in 1988 [17]. The matrix comprised cobalt-nanoparticles suspended in glycerol as the basic system into which the analyte was dissolved, similar to the sample preparation of FAB. Several other nano- and micro-particles were tested later, and results obtained that were comparable to those of Tanaka [18]. For his technique of surface-assisted laser desorption/ionization (SALDI), Sunner and coworkers used dry carbon and graphite substrates [19]. Another technique which has attracted much interest for the analysis of smaller molecules (and which is described in more detail in Chapter 9) was reported by Siuzdak [20]. This method, termed desorption/ionization on silicon (DIOS), uses preparations of neat organic samples on porous silicon. Several other methods and acronyms use similar systems, such as nanowires or sol–gel systems. All of these methods use the substrate on which the analyte is prepared for the absorption of the laser energy, and are characterized by a sensitivity which is lower than that of MALDI by several orders of magnitude, as well as a strongly increased ion fragmentation which limits the accessible mass range to somewhere between 2000 and 30 000 Da, depending on the method. There is reason to believe that all of these methods are based on a thermal desorption at the substrate/analyte interface, with the high internal excitation of the ions and low ion yield typical for thermal desorption processes. The very high heating and cooling rates, together with high peak temperatures of the substrates as well as the suspension of the absorbers in glycerol, apparently soften the desorption somewhat, the latter most probably through adiabatic cooling in the expanding plume; derivatization of the surfaces can up-concentrate the analyte of interest at the surfaces to increase the sensitivity. Indeed, a yoctomole (10−21 mole) sensitivity has been achieved in this way with a perfluorophenyl-derivatized DIOS system for a small hydrophobic peptide [21].
Recently, another technique termed matrix-assisted inlet ionization (MAII) was described by Trimpin et al., which enables the generation of multiply charged analyte ions similarly to those observed with ESI [22]. Although the analytes to ionize are also cocrystallized with typical MALDI matrices, the energy required for material ablation can be supplied by laser irradiation (laserspray ionization with laser pulse energies of about 10-fold that typically used for MALDI MS [22]), but is not limited to it [23]. It is assumed that analyte ion generation occurs independently of, and subsequent to, the ablation process in a heated inlet tube connecting the atmospheric pressure source to the vacuum of the mass spectrometer. Until now, the process of charge generation has not been completely understood; however, matrix evaporation of ablated highly charged clusters/droplets within the heated tube seems to explain the generation of multiply charged analytes. Due to the aforementioned differences to common MALDI MS, this technique is not discussed in this book.
What, then, is so special about the chemical matrix in MALDI? Some of its important features, such as the absorption of the laser energy, are easily understood, but rather surprisingly the overall process of the desorption and ionization has not yet been fully described, almost 30 years after its invention. Considerable progress regarding the mechanism of analyte desorption and protonation was recently achieved [24, 25]. Meanwhile, the search for better (i.e., more sensitive) matrices does not remain completely empirical, as some of the critical parameters for efficient analyte protonation (see Section 1.5) are uncovered, although other aspects such as prediction and targeted manipulation of the matrix morphology remain [26].
One important feature is the way in which the matrix and analyte interact in the MALDI sample. In a typical UV-MALDI “dried droplet” sample preparation, small volumes of an about 10−6–10−9 M solution of the analyte and a near-saturated (ca. 0.01–0.1 M) solution of the matrix are mixed; the solvent is then evaporated before the sample is introduced into the vacuum of the mass spectrometer. Upon solvent evaporation, the matrix crystallizes in a specific morphology to form a bed of small crystals that range in size from a few to a few hundred micrometers, depending on the matrix, the solvent, the substrate surface characteristics, and further details of the preparation. The typical molar analyte-to-matrix ratio ranges from about 10−2 for less-sensitive compounds such as poorly protonable drugs without basic functionalities, to approximately 10−7 for highly sensitive analytes, for example, quaternary ammonium-derivatives such as phosphatidylcholines or many basic peptides. The sample preparation is discussed in more detail in Section 1.8. One of the early surprises in MALDI development was that all of the well-functioning matrices known at that time incorporated the analytes in the crystals quantitatively (up to a maximum concentration), and in a homogeneous (on the light microscopic resolution level of 0.5 μm) distribution. This was shown for the UV-MALDI matrices 2,5-dihydroxybenzoic acid (2,5-DHB) [27], sinapic acid [28], α-cyano-4-hydroxycinnamic acid (HCCA) [29], and 3-hydroxypicolinic acid [29] as well as the IR-MALDI matrix succinic acid [30]. This homogeneous incorporation, in conjunction with the also homogeneous energy deposition and material ablation (for a discussion, see Section 1.3) resulted in the codesorption of intact nonvolatile and labile molecules with the matrix and, in addition, in a cooling of their internal energy in the expanding plume of material. Although the mechanisms and driving force for analyte incorporation are still largely unknown, attractive ion–ion interactions between dissolved protonated analytes and matrix acid anions during sample preparation seem to alleviate analyte inclusion/incorporation within the growing matrix crystals [26]. At typically slightly acidic sample preparation conditions, many analytes (such as most peptides and proteins) carry net positive charges due to protonated basic and predominantly neutralized acidic functionalities. Although matrix acid functionalities are also mostly neutralized under such pH conditions, the large molar matrix-to-analyte excess effects matrix anion amounts sufficiently high for at least stoichiometric analyte interactions in solution. These ion–ion interactions might provide the explanation of why almost all compounds used as matrices for the analysis of basic group containing analytes – which is the case for many natural drug classes – exhibit acid functionalities. Neutral α-cyanocinnamic acid (CCA) derivatives, for example, matrix amides or esters with weaker ion–dipole interactions between protonated analytes and neutral matrices, indeed resulted in strongly diminished analyte signal intensities.
Nevertheless, further prerequisites for successful analyte inclusion/incorporation presumably must be fulfilled. Krueger et al. found clear proof for homogeneous analyte incorporation by using colored pH indicators [31], whereas Horneffer et al. have shown in a systematic study of different position isomers of dihydroxybenzoic acids that only 2,5-DHB incorporates homogeneously and quantitatively, whereas other isomers such as 2,6-DHB do not incorporate at all, while some others incorporate only randomly [32]. Confocal laser scanning images of the protein avidin, labeled with the fluorochrome Texas Red for single crystals of 2,5-DHB and 2,6-DHB, are shown in Figure 1.1. No obvious correlation between analyte incorporation and the crystal structure of these isomers was found.
The state of the incorporated analyte molecules in the matrix crystals is another interesting question. Based on results obtained for the incorporation of pH-indicator dye molecules, Krueger et al. have concluded that molecules retain their solution charge state in the crystal, which implies that they also retain their solvation shell [31]. Horneffer et al. have found a high density of cavities of 10–2000 nm size in crystals of both 2,5-DHB and 2,6-DHB by electron microscopy [33]. At first sight, these cavities could be assumed to contain analyte molecules with residual solvent. However, if this is the case it is difficult to understand why 2,5-DHB – but not 2,6-DHB – incorporates analytes into these cavities; attempts to localize gold-labeled proteins in the cavities of 2,5-DHB were also inconclusive [33].
A solventless method for sample preparation was developed originally for the MALDI-MS of synthetic polymers, which often are not soluble in standard solvents [34]. In this method, matrix and analyte powders are mixed and ground in a mortar or ball-mill and then applied to a MALDI target support. It was shown that analyte spectra can be obtained from such preparations, even though the analyte is only chemisorbed at the matrix crystal surfaces [35]. However, the desorption is much less “soft” than MALDI-MS from samples with incorporated analytes, leading to a strongly increased metastable fragmentation of the ions and an upper mass limit for proteins of 30–55 kDa.
The role of the optical absorption of the matrix in the transfer of energy from the laser beam to the sample is governed by Beer's law [14]
(1.1)
where H is the laser fluence at depth z into the sample, H0 is the laser fluence at the sample surface, and α is the absorption coefficient (see Chapter 2, Section 2.2 for a definition of the fluence). The absorption coefficient α equals the product of the wavelength-dependent molar absorption coefficient αn which is a property of the matrix compound and the concentration cn of the absorbing molecules in the sample. The molar absorption coefficient αn has a maximum value for UV-MALDI irradiation of typical matrices between 5 × 103 and 5 × 104 l mol−1 cm−1 at the peak absorption wavelength. Molar absorption coefficients of this order of magnitude and at low wavelengths in the range of 300–400 nm are only provided by molecules with aromatic systems (typical matrix structures for instance contain phenyl or styryl derivatives) supported by electron-donating groups such as hydroxy residues. The exact wavelength of maximum absorption and its magnitude are determined by the position and nature of the ligands of the core ring, and are tabulated in a variety of reference sources. Some care should be exercised in using the tabulated values for αn, because they all refer to dilute solutions of the compounds. Compared to the absorption profiles of dissolved compounds, the absorption bands of MALDI samples in the solid state are typically broadened and slightly red- or blue-shifted in dependence on the strength of the chromophore–solvent interactions of the dissolved compounds [25]. The concentration cn of absorbers (chromophores) is unusually high in solid-state MALDI samples (about 10 mol l−1), taking into account the typical solid-state density of crystals of roughly 2 g cm−3 (e.g., 2,3-dihydroxybenzoic acid exhibits a density of 1.54 g cm−3), because all of the solvent is evaporated before the sample is introduced into the vacuum. As a result, the typical UV absorption coefficient α ranges from about 5 × 104 to 5 × 105 cm−1 at a laser wavelength of 337 or 355 nm. The inverse of α is called the penetration depth δ, and this has values of only 20 to 200 nm. It is the depth into the sample, at which the fluence has decreased to about 30% of the value at the surface. It is also an order of magnitude estimate of the depth of material ablated (desorbed) per single laser pulse in MALDI. Because of this very shallow ablation depth, a given location of the sample can usually be irradiated many times before the material is exhausted. For the MALDI process, the energy absorbed per unit volume Ea/V of the sample (loosely called “energy density”) is the process-determining quantity. This can be derived from Eq. (1.1) by simple differentiation to:
(1.2)
Equation (1.2) is at the core of the MALDI process. If a matrix is chosen with a sufficiently high absorption coefficient α, a relatively low fluence H0 suffices for achieving the critical “energy density” necessary to initialize ablation and ionization of a top layer of the sample. Values for H0 of 50–500 J m−2 are representative for most UV-MALDI applications.
As discussed in Chapter 2, Section 2.2, pulsed lasers with a pulse width of a few nanoseconds are employed in UV-MALDI. At a fluence of about 100 J m−2 and a pulse width of 2 ns, the “intensity” (irradiance) of the laser beam at the sample surface is only 1011 W m−2 or 107 W cm−2, which is not enough to induce any nonlinear absorption such as nonresonant two-photon absorption. For the linear absorption, the absorbed energy per unit volume can be controlled meticulously with a suitable variable attenuator in the laser beam, a feature which has emerged as being crucial for the successful MALDI of large molecules, because the desorption of nonvolatile, labile molecules can only be achieved in a narrow range of energy “density” between low-energy conditions insufficient for ablation and ionization and high-energy conditions leading to extensive analyte fragmentation (see Section 1.6). The other essential feature of this laser absorption is that the energy is transferred more or less uniformly to a macroscopic sample volume (except for the attenuation of the fluence into the sample and the fluence profile, as discussed in Chapter 2, Section 2.2). This is very different from the situation in SIMS or PD, where incident particles create minute tracks of atomic dimensions of very high energy density in the sample, with a strong radial decline of energy density. This strongly heterogeneous energy distribution is the main reason for the limitation of these methods for the intact desorption of larger molecules. The fluence can also be converted into a value for the photon flux – that is, the number of photons impinging on the sample per single laser pulse. A typical fluence of 100 J m−2 [36] corresponds to a photon flux of 1.7 × 1016 photons per cm2; each carrying an energy of 3.7 eV at the wavelength of 337 nm of the N2 laser. A molar absorption coefficient of 104 l mol−1 cm−1 represents a physical absorption cross-section of the chromophore of 1.6 × 10−17 cm2, resulting in an average of 0.3 photons absorbed per surface matrix molecule (about 110 kJ mol–1 matrix for 337 nm photons) for any given laser exposure at this fluence. For these considerations, it is assumed that the vast majority of electronic excitation energy is converted into lattice energy by internal conversion (as compared to processes such as fluorescence and chemical reactions). This is a very high density of excitation energy, close to the sum of all noncovalent bond energies of the ablated matrix volume. It is, therefore, not surprising that such a large amount of deposited energy leads to an explosive ablation of the excited sample volume. On the other hand, it renders even resonant two-photon absorption by the matrix rather unlikely. The high density of excited molecules does, however, result in a rather high rate of energy pooling in the sample, in which two neighboring electronically excited molecules pool their energy, with one of them acquiring twice the energy of the first excited singlet state (S1,v = 0) and the other falling back to its ground state [37]. This energy pooling is an important feature in some models for the ionization of the matrix molecules, which requires at least the energy of two photons for an initial photoionization of the matrix molecules [3, 38]. It is elucidated in more detail in Section 1.5.
The situation is similar, but not equal, for IR-MALDI. Optical absorption in the IR region of the spectrum represents a transition between vibrational and/or rotational molecular states. The probabilities for these transitions are typically lower than the electronic transitions in the UV by one to two orders of magnitude. The strongest such transitions are those of the O–H and N–H stretch vibrations near a 3 μm wavelength. The absorption coefficient of water or vacuum-stable ice, but also of the common IR-MALDI matrix glycerol, reaches peak values of 104 cm−1 in this wavelength region, corresponding to a penetration depth of about 1 μm, which is more than 20-fold that of typical penetration depths in the UV. As a result, the ablated mass per laser exposure in IR-MALDI exceeds that of UV-MALDI by at least a factor of 10, and the sample consumption rate is accordingly higher. Typical laser fluences for IR-MALDI range from 103 to 5 × 103 J m−2. Nonlinear absorption processes are even less likely for such fluences in the IR- as compared to UV-MALDI, and for the photon energy of only 0.4 eV or less even the absorption of several photons by a given chromophore or energy pooling cannot possibly excite single molecules to anywhere near their ionization energy.