7.1. SUMMARY

This chapter describes the basic technology and application of matrix-assisted laser desorption ionization (MALDI) imaging mass spectrometry (MS) in the field of neurobiology. Examples of applications that search for biomarkers of diseases of the central nervous system for both diagnostic and prognostic purposes are discussed. Newly developed strategies and analytical methods are discussed for the study of drug pharmacokinetics, neuropeptide metabolism, and treatment-induced or disease-specific effects on the proteome.

7.2. INTRODUCTION

The clinical diagnosis of many major neurodegenerative disorders remains unsatisfactory and consequently there is a need for biomarkers for both diagnostic and prognostic purposes. Moreover, as protocols and methods for high-throughput genomics and proteomics improve, these approaches will provide significant insight into the pathogenesis of many neurodegenerative diseases. The possibility of identifying novel biomarkers, those related to pathogenesis in particular, also promises to help achieve early detection and ultimately to help lead the way toward individualized treatment. Mass spectrometry is a molecular analysis technology that is vital to this biomarker discovery for proteomics.

7.3. MS TECHNOLOGY

MALDI MS is able to detect and measure levels of neuropeptides and proteins and their distributions directly in tissue sections. The discovery and description of MALDI MS was first reported in 1987 (1). Briefly, a laser beam (i.e., a nitrogen laser at 337 or Nd:Yag laser at 355 nm) irradiates a solid sample that is a co-crystal of sample molecules and molecules from an organic matrix material. Absorption of energy from the laser by the matrix causes an ablation of the irradiated surface of the sample, with the desorption and formation (for the most part) of singly protonated molecular species ([M+H]⁺). The desorbed ions are given an acceleration of 20–25 kV, traverse a time-of-flight (TOF) analyzer, and their mass-to-charge (m/z) is determined. Modern instruments have delayed extraction to achieve high sensitivity, a reflectron to record high-resolution mass spectra at low m/z values (<5000), and post-source decay or MS/MS capabilities to provide structure elucidation (e.g., TOF/ TOF or quadrupole TOF [QTOF] for peptide sequencing; for a detailed discussion of MS see Chapters 5 and 6).

The MALDI matrices generally employed for the analysis of peptides and proteins are sinapinic acid (3, 5-dimethoxy-4-hydroxycinnamic acid), CHCA (alpha-cyano-4-hydroxycinnamic acid), and DHBA (2, 5-dihydroxybenzoic acid). Matrix can be applied directly on tissue sections by a spray or spot deposition process to achieve a uniform coating so that the spatial relationships of compounds remain intact.

One of the most common mass analyzers used in combination with MALDI is the TOF mass analyzer, which is well suited for pulsed ion sources. The TOF mass analyzer measures the time it takes for an ion to travel a specific distance in a field-free flight tube to the detector. Typically 30–100 or more laser shots are averaged in order to obtain a spectrum. The sensitivity of commercially available MALDI instruments is in the attomole to low femtomole range, with a mass accuracy of 10–50 ppm using delayed extraction and a reflector to refocus the molecular ions in the mass analyzer. MALDI-TOF instruments are able to detect ions over 200 kDa, although sensitivity falls off rapidly with increased mass range. Other mass analyzers have also been coupled with a MALDI including quadrupole, ion trap, Fourier transform ion cyclotron resonance (FT-ICR), orbitrap, and QTOF (quadrupole TOF MS/MS instrument), as well as other analyzers (see Chapter 5 for a description of different types of MS instruments). A potentially useful instrument for imaging mass spectrometry studies is the FT-ICR mass spectrometer because it is able to measure the molecular weight of peptides and other low molecular weight compounds with an accuracy of 0.1–1 ppm.

We have utilized a MALDI-TOF mass spectrometer for the direct analysis of tissue for spatial detection of molecules using a laser spot size on target of 30–80 microns (2,3). This technology is termed “MALDI imaging mass spectrometry” and will be abbreviated “IMS” for the purposes of this chapter. The versatility of mass spectrometry-based proteomic approaches makes IMS an ideal analytical tool for many applications, including pharmacokinetic studies of centrally acting drugs (4–6), lipidomics (7), neuropeptide metabolism (8,9), and discovery-based proteomics (10–13).

7.3.1. Sample Preparation and Analysis

A thin section (typically 10–12 μm thick) of tissue is cut from a fresh frozen sample and thaw-mounted directly onto a MALDI target plate or compatible glass microscope slide (14). The section is covered with matrix dissolved in acidified organic solution (l%–2% triffuoroacetic acid in either 50% methanol or 50% acetonitrile in water) using a robotic spotter or spray coating device. Spotting robots such as the acoustic picoliter droplet ejector (Portrait 630; LabCyte) and a piezoelectric based chemical printer (ChIP-1000; Shimadzu) have been used with good results, including high sensitivity and highly reproducible spectra (15–17). Spray coating can be performed in several ways, for example, by hand using a spray nebulizer or with a commercially robotic device such as the ImagePrep (Bruker Daltonics) or TM nozzle sprayer (Leap Technology).

Irradiation of a spot on a tissue or biopsy section with 50–200 or more repetitive laser shots gives rise to a mass spectrum containing signals from molecules that were contained in that spot. Depending on the size of the sample and the analytical task at hand, many thousands of such mass spectra can be obtained from a single section through the use of a movable stage that allows repositioning of the area to be irradiated within milliseconds. In the profiling mode, the analytical task usually involves comparison of spectra from a relatively small number of spots on a sample either taken at random or targeted to specific areas in the tissue through histology or other means as shown in Figure 7.1a (44). Typically, a biopsy from a relatively homogeneous area of a tissue section would be interrogated in this way. In the imaging mode, a raster of the surface of the sample by the laser beam is accomplished by incremental movement of the sample stage to acquire thousands of spectra in an ordered array. These basic processes are illustrated in Figure 7.1b. The laser-ablated spots can be made adjacent (on 30–50 micron centers) for high-resolution imaging or at any other desired spacing. Often, a survey image is first obtained at 200–300 micron centers prior to a high-resolution image of a specific area of interest. The mass spectrum of each laser spot typically contains 500–1000 or more recorded ion species, with accurate mass assignments (± 1 Da in 10,000 up to about 30,000 Das) for most of these. Computer-generated images at a given molecular weight can be obtained by plotting the intensity value of the chosen molecular species in the ordered array of laser spots (or pixels). Special software has been implemented in our lab in order to speed and automate the imaging process. Spot-to-spot cycle times can be 1 sec or less, depending on the choice of user-set parameters and sample quality desired, enabling a 1000 pixel image to be acquired in less than 20 minutes. Even so, this is not a limiting speed and total acquisition times can be improved with the use of high-speed lasers and data transfer technology.

FIGURE 7.1

(a) Region-specific mass spectra can be obtained through histology-directed tissue profiling MS, in this example, targeting the cortex (blue) and striatum (red) in a rat brain tissue section. Two average mass spectra display different molecular signatures (more...)

The lower limit of imaging resolution is determined by (i) the matrix spot size and spot-to-spot distance in the case of robotic spotters, and (ii) by the size of the laser spot, currently approximately 30–50 μm, in the case of spray coating. Other important factors include tissue quality, the sample preparation protocols employed, section size, and user-determined time for analysis and data file size. In order to maximize reproducibility and sensitivity, in terms of peak height measurements and the number of molecular species detected, it is usually necessary to establish the optimum matrix-to-analyte ratio.

Each IMS image file contains multiple data dimensions: x and y coordinates for spatial localization, many hundreds of mass dimensions in terms of mass-to-charge (m/z), and the corresponding signal intensity at each m/z value. Using commercially available software or freeware such as the imaging software tool BioMap (Novartis), individual peptide or protein ion images can be visualized. Comparison of the relative intensities from normalized spectra then can be used to determine the relative amounts of a given compound throughout a section or series of samples.

7.3.2. Data Processing

Data pre-processing is generally necessary and includes baseline removal, noise reduction, and normalization. These are extremely important steps and validation of the procedures and algorithms used is highly recommended. In a recent study, several normalization algorithms were evaluated for their ability to reduce spectrum-to-spectrum variation (18). One of the most robust algorithms tested is based on the total ion current, i.e., the sum of all ion counts in a spectrum from the mass range 2–20 kDa. By rescaling each spectrum to its total ion current, unevenness in matrix application and crystallization is normalized, leading to increased signal-to-noise and more biologically relevant information (Figure 7.2). Preliminary data from our laboratories suggest this normalization strategy is also beneficial for the image analysis of low-molecular-weight species and peptides in the mass range of 400–3000 Da.

FIGURE 7.2

Baseline subtraction, noise removal, and total ion current normalization greatly improves image information. The three examples of myelin basic protein, thymosin beta-4, and PEP-19 were derived from a single IMS data file. Insert in the lower left panel (more...)

Data reduction for the purpose of comprehensive statistical analyses is an important step in biomarker discovery or exploratory proteomics. For MALDI IMS data, one generally accepted strategy is to apply a peak-picking algorithm for extraction of information regarding protein or peptide species present. This reduces the spectrum to a peak list containing one m/z value with its associated intensity per peak. Further data reduction can be done by averaging multiple spectra from a single organ or brain region that then is representative of that sub-region, although this must be done with great care so as not to obscure real spatial differences. In the end, the user must determine the trade-off of data file size, the speed of the image acquisition, and the resolution required since these parameters are inter-dependent.

7.4. MALDI IMAGING MASS SPECTROMETRY OF THE BRAIN

IMS allows determination of the tissue distribution of hundreds of peptides and proteins virtually at the same time. Since all of the images produced are inherently registered to each other, comparing multiple protein distributions from IMS images is straightforward (Figure 7.3). This is in contrast to dual- or multiple-antigen immunohistochemistry where each antibody must be optimized for each protein.

FIGURE 7.3

Rat brain sections at the level of the striatum were analyzed by MALDI IMS, and the distribution of three proteins were visualized using three primary colors: green, red, and blue. The spatial (lateral) resolution of these images is 300 µm.

An example of a mass spectrum that can be obtained directly from a tissue specimen in a profile experiment is shown in Figure 7.4 for a portion of the spectrum obtained from a human glioma biopsy. The complete spectrum is typically recorded over a mass-to-charge range of 2000–50,000, with most signals present in the lower half of this range. The vast majority of the signals in this spectrum are derived from proteins, verified through on-tissue protease digestion and subsequent sequence analysis of the resulting peptides (17). High-molecular-weight proteins (>200,000) can also be recorded directly from tissue, although this requires additional time for adjusting acquisition parameters to optimize sensitivity for this range.

FIGURE 7.4

(a) A human glioma specimen was mounted on a glass slide and a mass spectrum was obtained directly from the section in a profiling experiment (b) The mass spectrum is typically recorded over a mass-to-charge range of 2000–100,000, with most signals (more...)

We have imaged a coronal section of mouse brain in which glioma GL261 cancer cells had previously been injected in vivo in one hemisphere of the brain and subsequently a tumor had developed. After 15 days post-injection, the animal was sacrificed and brain sections were cut at 12-(μm thickness and mounted on a target plate. A tumor of several millimeters had developed in the brain left lateral ventricle. Evidence of metastatic migration of the tumor was seen in the right lateral ventricle. Significant differences in the profiles were observed, e.g., the signals at m/z 4965, 6719, 8565, and 12,134 were found consistently expressed at relatively higher levels in normal tissue, whereas m/z 9737, 9910, 11,641, and 12,372 were found consistently expressed at relatively higher levels in the tumor. A low-resolution image analysis of a section from this brain is shown in Figure 7.5. The perimeter of the tumor has been outlined in the optical image (panel a) for ease of viewing of the figure. Image analysis was performed with a resolution of 110 μm, averaging 20 laser shots per spectrum. Eight different protein expression maps across the section are shown in Figure 7.5b-i, taken from hundreds of such maps that could be generated from a single raster acquisition.

FIGURE 7.5

(a) A coronal section of mouse brain with a glioma tumor in one hemisphere; for ease of viewing the tumor is outlined with a dotted line. IMS analysis was carried out at 110-μm resolution and (b-i) display several molecular images, including histones (more...)

7.5. IMAGING MASS SPECTROMETRY OF NEUROPEPTIDES

One of the most challenging aspects of neurochemistry is the detection of endogenous neuropeptides due to their low in vivo concentrations ranging from pico- to sub-femtomolar levels (19,20). Recently, we have a developed a new, sensitive, and highly reproducible protocol for IMS of peptides and neuropeptides in rat brain sections. Utilizing a matrix spotter, approximately 1000 monoisotopic molecular species were observed and imaged in one section of a rat ventral midbrain (16). Furthermore, we have been able to identify neuropeptides such as substance P (SP) in the substantia nigra pars reticulata by MS/MS analysis directly off the rat brain tissue sections using a MALDI QTOF mass spectrometer. The capability of determining a neuropeptide sequence directly in a discrete brain region of a section enables the study of the physiological and disease-related metabolic processing of neuropeptides. For example, disease-state alterations in the metabolic processing of dynorphin A occurs in an animal model of Parkinson’s disease (PD) (21) that is well in line with findings of disturbed opioid transmission in both animal models and patients with Parkinson’s disease (22,23). Indeed, we have observed both full-length SP as well as the metabolites SP(l-7) and SP(l-9) in the rat substantia nigra pars reticulata (16).

7.6. MALDI PROFILING AND IMAGING OF EXPERIMENTAL NEURODEGENERATIVE DISORDERS

Animal models have been used extensively to better understand the etiology and pathophysiology of human neurodegenerative diseases. Ideally, they should reproduce the clinical manifestation of the disease and mimic the selective neuronal loss in the brain. These models can also be used to test therapeutic approaches for treating functional disturbances observed in the disease of interest.

Profiling and imaging MS technology has been applied to animal models of neurodegenerative disorders, particularly to compare peptide and protein patterns in different animal models of PD (13,24,25) and Alzheimer’s disease (AD) (3,26). Clinically, PD is characterized by motor features including resting tremor, bradykinesia, and rigidity, and AD by loss of mental functions (dementia). In these affected brains, peptide and protein patterns are complex and new molecular tools, such as MALDI IMS, are needed to identify not only the changes in these patterns but also to identify specific proteins involved in neurodegeneration.

Profiling MALDI MS has been used to generate peptide and protein profiles of brain tissue sections obtained from experimental PD (the unilaterally 6-hydroxydopamine lesioned rodent model). Initially, a number of protein expression profile differences were found in the dopamine denervated side of the brain in specific brain regions when compared to the corresponding intact side, for example, calmodulin, cytochrome c, and cytochrome c oxidase (13). This work was the first study utilizing MALDI MS protein profiling on tissue sections to screen for protein expression differences in an animal model of PD.

In addition to the loss of dopamine-producing neurons in substantia nigra, idiopathic PD is characterized by the presence of Lewy bodies. These are abnormal cytoplasmic protein inclusions found in the midbrain of PD patients and are regarded as a pathological characteristic of PD (27). A major component of Lewy bodies is ubiquitin (8.5 kD, 76 residue protein) (28). The ubiquitin pathway degrades intracellular proteins with a ubiquitin tag that marks proteins for proteolysis. The polyubiquinated proteins are enzymatically degraded to peptides, and the ubiquitin moieties are released intact (29). The high levels of ubiquitin and ubiquinated proteins in Lewy bodies suggest that protein degradation is impaired in PD. However, Lewy body formation has not been demonstrated in the acutely induced neurotoxic experimental models of PD (30). Using profiling MALDI MS on brain tissue sections from the unilaterally 6-OHDA lesioned rat model of PD, increased levels of free ubiquitin specifically in the dorsal striatum of the dopamine depleted hemisphere were detected (24). No similar changes were found in the intact hemisphere or in the ventral striatum of the dopamine depleted hemisphere. In addition, this study emphasizes the utility of MALDI MS for direct analysis of proteins and peptides on biological tissue sections. Most antibodies that have been used in immunocytochemical studies to assess the presence of ubiquitin in PD-associated Lewy bodies are directed toward a protein-bound form of ubiquitin, because free monomeric ubiquitin is not immunogenic in most mammalian species used to produce antibodies.

The 12 kDa neuroimmunophilin protein FKBP-12 is abundant in the brain and acts as a receptor for the immunosuppressant drug FK506 (31). The drug belongs to a class of neuroimmunophilin compounds that elicit both neuroprotection and neuroregeneration in neurotoxic chemical models of PD (32). Further, FK506 has also been shown to enhance nerve regeneration in a variety of experimental situations including nerve crush and transection, and spinal cord injury (33–35). In the presence of the drug FK506, FKBP-12 binds to a subunit of the calcium-dependent phosphatase calcineurin by blocking the access to the catalytic site of the subunit (36). Increased levels of FKBP-12 in the dorsal and middle part of the dopamine-depleted striatum were detected using the 6-OHDA model of PD (37). The finding was further confirmed by in situ hybridization, Western blotting, two-dimensional gel electrophoresis, and liquid chromatography electrospray tandem mass spectrometry. Another calcium-binding protein was effectively imaged from brain in the MPTP mouse model of PD (25). PEP-19 (Purkinje cell protein 4; PCP 4) is a 6.7 kDa protein that belongs to a family of proteins involved in calmodulin-dependent signal transduction (38). IMS showed that PEP-19 protein was predominantly localized to the striatum of the brain tissue cross sections and that PEP-19 levels were significantly reduced by 30% after MPTP administration (25) (Figure 7.6).

FIGURE 7.6

MALDI IMS analysis of brain tissue sections of an experimental model of PD. The relative ion density of PEP-19 from one control (a) and one MPTP-administered animal (b) displays a reduction of PEP-19 expression in the striatum (lateral resolution 280 (more...)

A pathological feature of AD is the presence of amyloid deposits in plaques and in blood vessel walls. These accumulations are mainly composed of amyloid beta (Aß) peptides of 4–5 kDa and are all derived from the amyloid precursor protein (APP) (39). In an animal model of AD, a transgenic mouse model over-expressing human APP23, the spatial distribution of different Aß peptides was studied using IMS (26,40). In this animal model of AD, an increased production of Aß and Aß peptide fragments has been shown to be induced by the mutation of the amyloid precursor protein gene, and A ß deposits develop in cortical and hippocampal structures upon aging. IMS detected the Aß peptides (1–37, 1–38, 1–39, 1–40, and 1–42) in the brain tissue sections located in the parietal and the occipital cortical lobe and in the hippocampus region. By normalizing the distributions to a standard protein insulin it was shown that Aß (1–40) and Aß (1–42) were the most abundant amyloid peptides (26,40).

7.7. THREE-DIMENSIONAL IMAGING MASS SPECTROMETRY OF THE BRAIN

The ability to spatially resolve different proteins while simultaneously acquiring quantitative information concerning their relative abundance is particularly important in heterogeneous tissues such as the brain. Analyses of three-dimensional (3D) brain data by different imaging modalities such as computer tomography, magnetic resonance imaging (MRI), and positron emission tomography (PET) have directly led to major advances in understanding both normal and pathological function. Most in vivo imaging methods provide only structural information, although PET can localize and quantitatively assess levels of certain receptors and transporters. Three-dimensional reconstructions of histological material, such as immunohistochemically stained tissue sections, can provide excellent anatomical resolution, but quantitation of such material is difficult and the use of antibodies to reveal the proteins of interest limits the approach to revealing few proteins in a given section and precludes any unbiased approach to monitoring proteins. By contrast, MALDI IMS detects hundreds of molecular species and recent developments have provided a novel basis of interrogation of protein localization in 3D (16,41,42). In one study, a series of ventral midbrain sections was collected and scanned (16). After MS acquisition, the matrix was washed off and the sections were stained using a histological stain (cresyl violet). Three-dimensional reconstructions of the brain region were made from the histologically stained sections, which then served as a template for inserting the registered MALDI IMS data into the volume reconstruction. This allowed the mapping of several known peptides and proteins, such as substance P and PEP-19, but also unknown proteins could be visualized simultaneously with a known protein (in this case PEP-19) and information gathered on the relative localization of the unknown species in the ventral midbrain (Figure 7.7).

FIGURE 7.7

Assessment of distribution in 3D of PEP-19 (green, top left panel) and an unknown 7.4 kDa protein (red, bottom left panel) in the rat ventral midbrain. Tissue sections 12-μm thick were collected at 200-μm spacing throughout the rat ventral (more...)

The structural heterogeneity of many tissues, such as brain, requires an approach like this that permits relatively high anatomical resolution coupled with quantitative analysis of proteins and the ability to do so in an unbiased fashion.

7.8. IMAGING WITH TANDEM MASS SPECTROMETRY

An extension of the early work using MALDI-TOF technology is the use of tandem mass spectrometry (MS² or MS/MS), whereby both the molecular species and also a key structural variety of the molecule are simultaneously monitored to give added validity to molecular specificity, particularly at low molecular weights. This involves an instrument containing two consecutive mass analyzers in conjunction with a collision-activated dissociation (CAD) process (see Chapter 5 for a discussion on fundamentals and instrumentation of mass spectrometry). A parent molecular ion is selected by the first mass analyzer and transmitted to a collision cell where it is fragmented by collisions with a neutral gas, such as helium or nitrogen. The fragments are then analyzed by the second mass analyzer. From the fragmentation pattern, the identity of the precursor ion can be determined. The use of MS/MS for image analysis of low-molecular weight compounds provides extremely high specificity and typically reduces levels of chemical noise, thus enhancing the signal-to-noise ratio and the limits of detection (6,43). The MS/MS imaging analysis is practically limited to a discrete number of parent ions that can be recorded in a given experiment. For example, Khatib-Shahidi et al. (4) demonstrated the distribution of clozapine and three of its metabolites at various time points after administration. Clozapine was clearly detected in the central nervous system, whereas the metabolites were mostly present at the sites of conversion: the liver and kidney.

7.9. QUANTITATION

The reproducibility and ability of the technology to provide relative quantitation have been published and cited in several reviews (44,45). Protocols have been developed for preparation of the section and matrix application that provide a high level of reproducibility. For example, in one study of glioma biopsies designed to test the reproducibility of the spectral patterns on a given biopsy from areas of similar histology, 687 individual spectra were obtained from 89 biopsies taken from 58 patients. Histologically, the samples were classified as 18 non-tumor, 25 grade 2, 17 grade 3, and 38 grade 4 biopsies. The statistical result showed 7% variability in the MS signal variation within a morphologically similar region of a given biopsy (46). A second issue is the question of reproducibility of a given signal from different parts of a tissue section where there are morphologically dissimilar areas. Although we have not performed a statistical analysis, at least in the case of a brain section, we do not observe significant variations for proteins known to be evenly distributed in tissue. Thus, for a coronal brain section, the image intensity of such a protein is quite similar throughout the section. However, MALDI MS of sections from different organs display significant variations in the spectral signature, reflecting in part the inherent complexity of the molecular compositions of the different cell types present.

7.10. IDENTIFICATION OF PROTEINS

Since the measurement of the molecular weight of a protein does not specifically identify the protein, additional studies are necessary and this can involve two approaches. First, protein extracts from tissue samples can be fractionated by high performance liquid chromatography (HPLC) (see Chapter 3), one-dimensional Polyacrylamide gel electrophoresis with sodium dodecyl sulfate (1D SDS-PAGE), or solution phase isoelectric focusing, and fractions having the protein of interest, as determined by MALDI MS, are then subjected to tryptic digestion and analysis by LC-MS/MS (as described in Chapter 5). Peptides and their corresponding proteins of origin are identified from MS-MS spectra with a protein database search program that correlates uninterpreted MS-MS spectra with theoretical spectra from database sequences. Confirmation of protein identities are based on the apparent molecular weight of the MS-MS identified proteins compared to pattern-specific signals detected in the MALDI profiles.

For MALDI IMS analysis, targeted proteins can be identified directly from tissue through on-tissue proteolysis. Trypsin is deposited using a robotic spotter in a Cartesian array across the tissue section. Multiple passes are required to keep the trypsin in solution and allow sufficient time for the enzymatic reaction to take place. Subsequently, matrix solution (e.g., DHB in 50% methanol, 0.3% TFA) is placed directly on top of the trypsin digests and these spots are analyzed by MALDI-TOF. The data set can be used to map tryptic peptides with similar distribution patterns using cluster analysis algorithms for first-stage peptide mapping analysis. Second, MS/MS analysis can be performed directly off tissue, which has the advantage of allowing the tryptic peptide parent mass and fragment masses to be mapped to a specific localization in the tissue (16,17). MS/MS analysis results in a list of tryptic masses coupled with a positive identification of a single protein. These tryptic masses can then be visualized in a section, mapping to the individual digest spots. Taken together, this in situ peptide analysis approach provides high power of molecular identification including spatial specificity. Another advantage of this approach is the possibility to map high-molecular-weight proteins that are otherwise not readily detectable in their intact form.

A second identification approach uses the combination of LC-MS-MS analyses and stable isotope tags. In its simplest form, equal amounts of protein extracts from two samples to be compared are chemically tagged with light and heavy (unlabeled and deuterium or ¹³C-labeled) reagents. They are combined and digested and the tagged peptides are analyzed by LC-MS-MS. Peptides derived from the two samples are distinguished by pairs of signals in full scan MS separated by the mass difference of the light and heavy isotope tags. Pairs of signals whose intensities deviate from unity represent proteins that were differentially present in the original two samples. MS-MS spectra acquired in the same LC-MS-MS analyses allow unambiguous identification of the differentially expressed proteins. Two commercial products are currently available. One commonly used procedure utilizes thiol-reactive isotope coded affinity tag (ICAT) reagents (47), and isobaric tags for relative and absolute quantification (iTRAQ), another acid-cleavable reagent that offers more efficient recovery of tagged peptides and produces high quality MS-MS spectra for identification (48). N-terminal isotope tagging of tryptic peptides enables identification of proteins that differ in posttranslational modifications rather than protein expression level per se (49,50). A major advantage of the tagging reagents is that they provide a quantitative comparison of the proteins of interest.

7.11. INTEGRATING HISTOLOGY AND IMAGING MASS SPECTROMETRY

An important aspect of IMS is the ability to identify histological features of interest within the tissue sections and then obtain ion images of that same area. A protocol has been developed that allows the histology and MS analyses to be performed on the same section (44), avoiding significant problems of comparing the histology and mass spectra taken from different sections. A conductive glass slide is used to mount tissue sections, allowing visualization of tissue sections with transmitted light and also serving as the target plate for IMS measurements. Typically, a pathologist studies a photomicrograph of the stained section and marks the areas of interest. In this way, the number of spectra are reduced to include only the tissue regions of interest, and a high number of specimens can be examined in a short time. Several tissue staining procedures compatible with IMS can be used to permit a pathologist to perform routine histology (hematoxylin and eosin [H&E] stained tissue produces poor mass spectral patterns). Figure 7.4 presents the analysis of a grade IV human glioma section after staining with methylene blue, providing a strong nuclear as well as faint cytoplasmic staining. The spot size on the tissue analyzed can vary from 80 μm up to 1 mm or larger. The resulting spectrum displays signals in the m/z range from 2000 to 70,000. Several other dyes may also be used, including toluidine blue, cresyl violet, and nuclear fast red (14).

Histology-directed profiling is especially important in the research of invasive tumors where connective tissues and normal intact cells often surround discrete clusters of cancer cells. Cornett et al. (44) demonstrated how this approach could discern, by unsupervised cluster analysis, different grades of breast tumor stages. Earlier forms of this strategy involved manually depositing matrix and subsequent manual acquisition of MS, which was successfully used for the classification of various glioma tumor grades and for the development of prognostic markers of short-term versus long-term survival of glioma patients (46).

7.12. CELL ISOLATION BY LASER CAPTURE MICRODISSECTION AND ANALYSIS BY IMS

Laser capture microdissection (LCM) is a technology that permits the isolation of single cells or single populations of cells from thin tissue sections, typically 5–10 μm thick, mounted on a glass slide (51). Using IMS, protein signatures can be obtained from 50–300 captured cells from a single cell type within a heterogeneous sample (52). For the Pixcell LCM instrument, a narrow laser beam (7.5–30 μm in diameter) irradiates a heat-sensitive polymer film on a cap that is in contact with a tissue section. The heated polymer adheres to the targeted cell(s) and these cells are subsequently removed from the section when the cap is lifted. We have established protocols to analyze cancer cells by MALDI MS after their LCM capture isolation from a thin tissue section (53).

7.13. LIMITATIONS OF TECHNOLOGY

IMS has several technical limitations, some of which are inherent in the laser desorption/ionization process and others in the performance of instrumentation available today, but advances are under way through which significant improvements will be made. Certainly one of the inherent limitations is the “ion suppression” effect that occurs in the MALDI ionization process. Basically, proteins that can act as better bases to capture protons can show preferential ionization yields. Although careful preparation protocols can minimize this effect, it can be significant. In that light, while one can estimate relative intensities for a given protein in several samples, it does not necessarily apply for two different proteins in a spectrum. Another limitation is the fact that only a small window of many hundreds of proteins are measured from a single spot on a sample, whereas many times this number exist in the tissue at moderate levels. Also, using current protocols, soluble proteins are favored since generally the preparation does not involve the use of detergents since these significantly compromise sensitivity. To this end we have synthesized several new families of cleavable detergents that, for example, when matrix is added to give low pH, the detergent hydrolyzes and does not interfere with the analysis (54). One of these detergents is equivalent to SDS in its detergent properties.

A practical limitation of IMS is the laser spot size and the trade-off between image resolution and sensitivity; the smaller the spot size the less material there is to desorb. Smaller spot sizes have been achieved and we have developed a 1–5 μm spot diameter on target, but its effectiveness in tissue analysis is yet to be established. Overall, the sensitivity of IMS on tissue is currently estimated to be in the high attomole range, but it is difficult to measure this accurately. Thus, proteins that are expressed at low copy numbers per cell will not be readily analyzed with current instrumentation.

Although the spatial resolution of MALDI IMS is relatively high, at this point in its development it is insufficient to accurately differentiate subcellular regions in single mammalian cells (such as axons and dendrites). Future advances in MALDI IMS suggest that it will be possible to localize peptides and proteins in much smaller areas (55). Ultimately, combining mapping of a given protein at high spatial resolution coupled with markers of cellular components (e.g., MAP2 for dendrites) should permit unambiguous determination of molecular entities in subcellular structures.

7.14. PERSPECTIVES

Neuropeptides are activated from preprohormones through specific proteases and subjected to posttranslational modifications (PTMs) prior to release. In general, over 200 posttranslational modifications have been reported to date, including glycosylation, C-terminal amidation, N-terminal acetylation, phosphorylation, and sulfation (56). Mass-spectrometry-based approaches are well suited for the detection and identification of previously unknown PTMs. Many PTMs depend on cell- and site-specific enzymes, which require a high spatial resolution in order to be detected. MALDI IMS is able to detect femtomolar levels of neuropeptides in tissue areas as small as 50 microns. Future work will aim to correlate PTMs and protein changes in specific brain areas using MALDI IMS, and promises to elucidate biochemical processes in both normal and pathological states.

Analysis of brain tissue sections using IMS offers many advantages over traditional analytical techniques such as Western immunoblotting, in situ hybridization histochemistry, and immunohistochemistry, which have to rely on known chemistry and well-characterized peptide and protein species. In contrast, IMS has the capability of detecting virtually any compound present in a tissue section, as long as it can be ionized and desorbed into a gaseous phase in the MALDI process. In addition, while proteolytic cleavage and PTMs of proteins and neuropeptides are common mechanisms for the fine-tuning regulation of biological activity, these events may be undetectable using immunology-based methods that target specific antigenic epitopes, but can be readily determined by MS-based methods, including IMS.

Many classical technologies and approaches for protein analyses, such as two-dimensional difference gel electrophoresis (2D-DIGE), are not commonly used for neuropeptide analyses for several reasons. This is because relatively low concentrations of neuropeptides make it difficult to obtain enough material from small regions or subregions of the target brain structure, and the mass range of neuropeptides is usually below that of the effective limits of gel-based approaches (i.e., ~5–10 kDa) (see Chapter 4 for a discussion of 2D-DIGE). In addition, the throughput for multiple samples is low. IMS provides strength in these areas because of its high sensitivity in the mass range up to 20 kDa and its extraordinary high-throughput capabilities. All in all, IMS is ideally suited for unbiased spatial localization and relative quantitation of proteins in brain and other heterogeneous tissues.

ACKNOWLEDGMENTS

The authors thank D. S. Cornett for excellent technical assistance and for the custom imaging preprocessing software, and Ariel Deutch for his consultation and constructive comments, and also acknowledge funding from NIH grant 2R01 GM58008–09, and the Department of Defense grant W81XWH-05–1–0179.

REFERENCES

1.

Karas M., Bachmann D., Bahr U., Hillenkamp F. Matrix-assisted ultraviolet-laser desorption of nonvolatile compounds. International Journal of Mass Spectrometry and Ion Processes. 1987;78:53–68.

2.

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;69:4751–60. [PubMed: 9406525]

3.

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;7:493–6. [PubMed: 11283679]

4.

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;78:6448–56. [PubMed: 16970320]

5.

Reyzer M. L., Caprioli R. M. MALDI mass spectrometry for direct tissue analysis: A new tool for biomarker discovery. J Proteome Res. 2005;4:1138–42. [PubMed: 16083264]

6.

Reyzer M. L., Hsieh Y., Ng K., Korfmacher W. A., Caprioli R. M. Direct analysis of drug candidates in tissue by matrix-assisted laser desorption/ionization mass spectrometry. J Mass Spectrom. 2003;38:1081–92. [PubMed: 14595858]

7.

McLean J. A., Ridenour W. B., Caprioli R. M. Profiling and imaging of tissues by imaging ion mobility-mass spectrometry. J Mass Spectrom. 2007;42:1099–1105. [PubMed: 17621390]

8.

Andren P. E., Caprioli R. M. Determination of extracellular release of neurotensin in discrete rat brain regions utilizing in vivo microdialysis/electrospray mass spectrometry. Brain Res. 1999;845:123–9. [PubMed: 10536191]

9.

Nydahl K. S., Pierson J., Nyberg F., Caprioli R. M., Andren P. E. In vivo processing of LVV-hemorphin-7 in rat brain and blood utilizing microdialysis combined with electrospray mass spectrometry. Rapid Commun Mass Spectrom. 2003;17:838–44. [PubMed: 12672139]

10.

Kruse R., Sweedler J. V. Spatial profiling invertebrate ganglia using MALDI MS. J Am Soc Mass Spectrom. 2003;14:752–9. [PubMed: 12837597]

11.

Lemaire R., Menguellet S. A., Stauber J., et al. Specific MALDI imaging and profiling for biomarker hunting and validation: Fragment of the 11S proteasome activator complex, Reg alpha fragment, is a new potential ovary cancer biomarker. J Proteome Res. 2007;6:4127–34. [PubMed: 17939699]

12.

Meistermann H., Norris J. L., Aerni H. R., et al. Biomarker discovery by imaging mass spectrometry: Transthyretin is a biomarker for gentamicin-induced nephrotoxicity in rat. Mol Cell Proteomics. 2006;5:1876–86. [PubMed: 16705188]

13.

Pierson J., Norris J. L., Aerni H. R., et al. Molecular profiling of experimental Parkinson’s disease: Direct analysis of peptides and proteins on brain tissue sections by MALDI mass spectrometry. J Proteome Res. 2004;3:289–95. [PubMed: 15113106]

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. [PubMed: 12898649]

15.

Aerni H. R., Cornett D. S., Caprioli R. M. Automated acoustic matrix deposition for MALDI sample preparation. Anal Chem. 2006;78:827–34. [PubMed: 16448057]

16.

Andersson M., Groseclose M. R., Deutch A. Y., Caprioli R. M. Imaging mass spectrometry of proteins and peptides: 3D volume reconstruction. Nat Methods. 2008;5:101–8. [PubMed: 18165806]

17.

Groseclose M. R., Andersson M., Hardesty W. M., Caprioli R. M. Identification of proteins directly from tissue: In situ tryptic digestions coupled with imaging mass spectrometry. J Mass Spectrom. 2007;42:254–62. [PubMed: 17230433]

18.

Norris J. L., Cornett D. S., Mobley J. A., et al. Processing MALDI mass spectra to improve mass spectral direct tissue analysis. Int J Mass Spectrom. 2007;260:212–21. [PMC free article: PMC1885223] [PubMed: 17541451]

19.

Kendrick K. M. Microdialysis measurement of in vivo neuropeptide release. J Neurosci Methods. 1990;34:35–46. [PubMed: 2259245]

20.

Strand F. L. Neuropeptides: General characteristics and neuropharmaceutical potential in treating CNS disorders. Prog Drug Res. 2003;61:1–37. [PubMed: 14674607]

21.

Klintenberg R., Andren P. E. Altered extracellular striatal in vivo biotransformation of the opioid neuropeptide dynorphin A(1-17) in the unilateral 6-OHDA rat model of Parkinson’s disease. J Mass Spectrom. 2005;40:261–70. [PubMed: 15706626]

22.

Johansson P. A., Andersson M., Andersson K. E., Cenci M. A. Alterations in cortical and basal ganglia levels of opioid receptor binding in a rat model of l-DOPA-induced dyskinesia. Neurobiol Dis. 2001;8:220–39. [PubMed: 11300719]

23.

Nisbet A. P., Foster O. J., Kingsbury A., et al. Preproenkephalin and preprotachykinin messenger RNA expression in normal human basal ganglia and in Parkinson’s disease. Neuroscience. 1995;66:361–76. [PubMed: 7477878]

24.

Pierson J., Svenningsson P., Caprioli R. M., Andren P. E. Increased levels of ubiquitin in the 6-OHDA-lesioned striatum of rats. J Proteome Res. 2005;4:223–6. [PubMed: 15822896]

25.

Skold K., Svensson M., Nilsson A., et al. Decreased striatal levels of PEP-19 following MPTP lesion in the mouse. J Proteome Res. 2006;5:262–9. [PubMed: 16457591]

26.

Stoeckli M., Knochenmuss R., McCombie G., et al. MALDI MS imaging of amyloid. Methods Enzymol. 2006;412:94–106. [PubMed: 17046654]

27.

Gibb W. R. Neuropathology in movement disorders. J Neurol Neurosurg Psychiatry. 1989 Suppl:55–67 [PMC free article: PMC1033309] [PubMed: 2547027]

28.

Fornai F., Lenzi P., Gesi M., et al. Recent knowledge on molecular components of Lewy bodies discloses future therapeutic strategies in Parkinson’s disease. Curr Drug Targets CNS Neurol Disord. 2003;2:149–52. [PubMed: 12769795]

29.

Snyder H., Wolozin B. Pathological proteins in Parkinson’s disease: Focus on the proteasome. J Mol Neurosci. 2004;24:425–42. [PubMed: 15655264]

30.

Melrose H. L., Lincoln S. J., Tyndall G. M., Farrer M. J. Parkinson’s disease: A rethink of rodent models. Exp Brain Res. 2006;173:196–204. [PubMed: 16639500]

31.

Liu J., Farmer J. D. Jr., Lane W. S., et al. Calcineurin is a common target of cyclophilin-cyclosporin A and FKBP-FK506 complexes. Cell. 1991;66:807–15. [PubMed: 1715244]

32.

Costantini L. C., Chaturvedi P., Armistead D. M., et al. A novel immunophilin ligand: Distinct branching effects on dopaminergic neurons in culture and neurotrophic actions after oral administration in an animal model of Parkinson’s disease. Neurobiol Dis. 1998;5:97–106. [PubMed: 9746907]

33.

Bavetta S., Hamlyn P. J., Burnstock G., Lieberman A. R., Anderson P. N. The effects of FK506 on dorsal column axons following spinal cord injury in adult rats: neuroprotection and local regeneration. Exp Neurol. 1999;158:382–93. [PubMed: 10415144]

34.

Gold B. G., Katoh K., Storm-Dickerson T. The immunosuppressant FK506 increases the rate of axonal regeneration in rat sciatic nerve. J Neurosci. 1995;15:7509–16. [PMC free article: PMC6578050] [PubMed: 7472502]

35.

Madsen J. R., MacDonald P., Irwin N., et al. Tacrolimus (FK506) increases neuronal expression of GAP-43 and improves functional recovery after spinal cord injury in rats. Exp Neurol. 1998;154:673–83. [PubMed: 9878202]

36.

Hemenway C. S., Heitman J. Calcineurin. Structure, function, and inhibition. Cell Biochem Biophys. 1999;30:115–51. [PubMed: 10099825]

37.

Nilsson A., Skold K., Sjogren B., et al. Increased striatal mRNA and protein levels of the immunophilin FKBP-12 in experimental Parkinson’s disease and identification of FKBP-12-binding proteins. J Proteome Res. 2007;6:3952–61. [PubMed: 17877381]

38.

Slemmon J. R., Feng B., Erhardt J. A. Small proteins that modulate calmodulin-dependent signal transduction: Effects of PEP-19, neuromodulin, and neurogranin on enzyme activation and cellular homeostasis. Mol Neurobiol. 2000;22:99–113. [PubMed: 11414283]

39.

Davison A. N. Pathophysiology of ageing brain. Gerontology. 1987;33:129–35. [PubMed: 3308642]

40.

Rohner T. C., Staab D., Stoeckli M. MALDI mass spectrometric imaging of biological tissue sections. Mech Ageing Dev. 2005;126:177–85. [PubMed: 15610777]

41.

Crecelius A. C., Cornett D. S., Caprioli R. M., et al. Three-dimensional visualization of protein expression in mouse brain structures using imaging mass spectrometry. J Am Soc Mass Spectrom. 2005;16:1093–9. [PubMed: 15923124]

42.

Sinha T. K., Khatib-Shahidi S., Yankeelov T. E., et al. Integrating spatially resolved three-dimensional MALDI IMS with in vivo magnetic resonance imaging. Nat Methods. 2008;5:57–9. [PMC free article: PMC2649801] [PubMed: 18084298]

43.

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. [PubMed: 17185024]

44.

Cornett D. S., Mobley J. A., Dias E. C., et al. A novel histology-directed strategy for MALDI-MS tissue profiling that improves throughput and cellular specificity in human breast cancer. Mol Cell Proteomics. 2006;5:1975–83. [PubMed: 16849436]

45.

Garden R. W., Sweedler J. V. Heterogeneity within MALDI samples as revealed by mass spectrometric imaging. Anal Chem. 2000;72:30–6. [PubMed: 10655631]

46.

Schwartz S. A., Weil R. J., Thompson R. C., et al. Proteomic-based prognosis of brain tumor patients using direct-tissue matrix-assisted laser desorption ionization mass spectrometry. Cancer Res. 2005;65:7674–81. [PubMed: 16140934]

47.

Gygi S. P., Rist B., Gerber S. A., et al. Quantitative analysis of complex protein mixtures using isotope-coded affinity tags. Nat Biotechnol. 1999;17:994–9. [PubMed: 10504701]

48.

Ross P. L., Huang Y. N., Marchese J. N., et al. Multiplexed protein quantitation in Saccharomyces cerevisiae using amine-reactive isobaric tagging reagents. Mol Cell Proteomics. 2004;3:1154–69. [PubMed: 15385600]

49.

Pflieger D., Junger M. A., Muller M., et al. Quantitative proteomic analysis of protein complexes: Concurrent identification of interactors and their state of phosphorylation. Mol Cell Proteomics. 2008;7:326–46. [PubMed: 17956857]

50.

Wiese S., Reidegeld K. A., Meyer H. E., Warscheid B. Protein labeling by iTRAQ: A new tool for quantitative mass spectrometry in proteome research. Proteomics. 2007;7:340–50. [PubMed: 17177251]

51.

Emmert-Buck M. R., Bonner R. F., Smith P. D., et al. Laser capture microdissection. Science. 1996;274:998–1001. [PubMed: 8875945]

52.

Xu B. J., Caprioli R. M., Sanders M. E., Jensen R. A. Direct analysis of laser capture microdissected cells by MALDI mass spectrometry. J Am Soc Mass Spectrom. 2002;13:1292–7. [PubMed: 12443019]

53.

Xu B. J., Shyr Y., Liang X., et al. Proteomic patterns and prediction of glomerulosclerosis and its mechanisms. J Am Soc Nephrol. 2005;16:2967–75. [PubMed: 16079267]

54.

Norris J. L., Porter N. A., Caprioli R. M. Mass spectrometry of intracellular and membrane proteins using cleavable detergents. Anal Chem. 2003;75:6642–7. [PubMed: 14640740]

55.

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;42:476–89. [PubMed: 17328093]

56.

Mann M., Jensen O. N. Proteomic analysis of post-translational modifications. Nat Biotechnol. 2003;21:255–61. [PubMed: 12610572]