Journal of Neuroimmune Pharmacology

, Volume 5, Issue 1, pp 31–43

Imaging Mass Spectrometry for Visualization of Drug and Endogenous Metabolite Distribution: Toward In Situ Pharmacometabolomes

Authors

  • Yuki Sugiura
    • Department of Bioscience and BiotechnologyTokyo Institute of Technology
    • Mitsubishi Kagaku Institute of Life Sciences
    • Department of Molecular AnatomyHamamatsu University School of Medicine
    • Mitsubishi Kagaku Institute of Life Sciences
    • Department of Molecular AnatomyHamamatsu University School of Medicine
Invited Review

DOI: 10.1007/s11481-009-9162-6

Cite this article as:
Sugiura, Y. & Setou, M. J Neuroimmune Pharmacol (2010) 5: 31. doi:10.1007/s11481-009-9162-6

Abstract

It is important to determine how a candidate drug is distributed and metabolized within the body in early phase of drug discovery. Recently, matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry (IMS; also referred to as mass spectrometry imaging) has attracted great interest for monitoring drug delivery and metabolism. Since this emerging technique enables simultaneous imaging of many types of metabolite molecules, MALDI-IMS can visualize and distinguish the parent drug and its metabolites. As another important advantage, changes in endogenous metabolites in response to drug administration can be mapped and evaluated in tissue sections. In this review, we discuss the capabilities of current IMS techniques for imaging metabolite molecules and summarize representative studies on imaging of both endogenous and exogenous metabolites. In addition, current limitations and problems with the technique are discussed, and reports of progress toward solving these problems are summarized. With this new tool, the pharmacological research community can begin to map the in situ pharmacometabolome.

Keywords

imaging mass spectrometryMALDIpharmacometabolome

Introduction

A large number of biomolecules ranging from small-metabolite molecules to much larger proteins have been visualized by the matrix-assisted laser desorption/ionization (MALDI) imaging mass spectrometry (IMS) technique in cells and tissues. This emerging imaging technique was initially developed as a tool for protein imaging (Caprioli et al. 1997; Chaurand et al. 1999; Stoeckli et al. 2001), and, so far, most reports on MALDI-IMS describe the detection and imaging of proteins or peptides (Chaurand et al. 2006; Stoeckli et al. 2002). On the other hand, research directed toward detecting and imaging small organic molecules has recently been expanding. Figure 1 shows the result of a PubMed search (reviews were excluded) using “imaging mass spectrometry” as key words. Reports are subdivided into groups according to the materials analyzed in the study, and the number of reports in each group is indicated. Notably, the number of reports regarding the IMS of small compounds has gradually increased and made up half of the publications in 2007.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig1_HTML.gif
Fig. 1

PubMed search result using “imaging mass spectrometry” as the keyword (partially adapted from Sugiura et al. 2008)

IMS is an effective technique for imaging the distribution of small metabolites, both exogenous drugs and endogenous metabolic intermediates, including lipids, amino acids, and organic acids. The emergence of IMS as a tool for metabolite imaging has an impact because we do not have an established technology for imaging metabolites, except for localization of transcripts by in situ hybridization with oligonucleotide probes, and localization of proteins using immunohistochemistry based on antibodies (Table 1). Due to the mass spectrometry (MS)-based detection principle, IMS has several advantages as a small-metabolite-imaging tool in tissues: first, IMS does not require any labels or specific probes; second, IMS is a nontargeted imaging method, so we can localize unexpected metabolites; finally, simultaneous imaging of many types of metabolite molecules is possible. Given the enormous molecular diversity of metabolite species, all these features are necessary for assessing metabolite localization. In addition, using tandem-MS analysis, the detailed structure of metabolite molecules can be identified directly on tissue sections (Garrett et al. 2006; Shimma et al. 2008). This technique is also effective for discriminating isobaric ions (Garrett et al. 2006; Khatib-Shahidi et al. 2006).
Table 1

Potential contribution of IMS for imaging of metabolites in tissues or cells

https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Tab1_HTML.gif

Workflow of imaging mass spectrometry

The workflow of MALDI-IMS is shown in Fig. 2. Basically, researchers take thin tissue slices mounted on conductive indium tin oxide (ITO) glass slides and apply a suitable MALDI matrix to the tissue section. Next, the ITO slide is inserted into a mass spectrometer. A focused laser beam is directed at predetermined positions in the tissue slice and the mass spectrometer records the spatial distribution of molecular species (typically with 10–200-μm scan pitch). Automated data collection takes 2–6 h, depending on the number of points assayed. Suitable image processing software can be used to import data from the mass spectrometer to allow visualization and comparison with the histological image of the sample.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig2_HTML.gif
Fig. 2

Schematic representation of MALDI-IMS procedures. Tissue section mounted on an ITO glass slide is covered with a specific MALDI matrix. Next, the ITO slide is inserted into a mass spectrometer. The MALDI laser scans through a set of preselected locations in the tissue (10–200 μm scan pitch) and the mass spectrometer records the spatial distribution of molecular species. Suitable image processing software can be used to import data from the mass spectrometer to allow visualization and comparison with the histological image of the sample. In this figure, green and red pixels represent the ion intensity of mass peaks labeled with asterisk and double dagger, respectively

Current IMS for small organic compounds

Today, the application of IMS to small organic compounds (m/z < 1,000) can be subdivided into two distinct areas: (1) measurement of exogenous drugs and (2) measurement of endogenous metabolites.
  1. 1.
    Mapping the metabolism of an administered drug and the distribution of its metabolites by IMS is attracting much attention because of its advantages over conventional imaging techniques, as will be discussed below. In this area, a number of studies on antipsychotics, cancer drugs, antianxiety drugs, and hypnotics have been reported (summarized in Table 2).
    Table 2

    Imaging/detection of small metabolite molecules in tissues by MALDI and related ionization techniques

     

    Endogenous metabolites

    Exogenous drugs

    Complex lipid

    Glycerophospholipids

     

     PCs (Astigarraga et al. 2008; Garrett et al. 2006; Hayasaka et al. 2008a; Jackson et al. 2005a, b)

     

     PEs (Astigarraga et al. 2008; Jackson et al. 2005a, 2007b)

     

     PIs (Astigarraga et al. 2008; Jackson et al. 2005a, 2007b)

     

     PSs (Astigarraga et al. 2008; Jackson et al. 2005a, 2007b)

     

     PGs (Jackson et al. 2005a, 2007b)

     

     Cardiolipins (Wang et al. 2007)

     

    Glycosphingolipids

    Olanzapine (antipsychotic; Cornett et al. 2008; Khatib-Shahidi et al. 2006)

     Gangliosides (Chen et al. 2008; Sugiura et al. 2008)

    Imatinib (cancer drug; Cornett et al. 2008)

     Sulfatides (Ageta et al. 2008; Chen et al. 2008; Jackson et al. 2007b)

    Vinblastine (cancer drug; Trim et al. 2008)

     Galactosyl-ceramide (Cha and Yeung 2007; Taira et al. 2008)

    Banoxantrone (cancer drug; Atkinson et al. 2007)

    Simple lipid

    Neutral lipids

    Diazepam (Kokaji 2008) and temazepam (antianxiety drugs, hypnotic; Kokaji 2008)

     Triacylglycerols (Astigarraga et al. 2008)

     

     Diacylglycerols (Astigarraga et al. 2008)

     

     Cholesterol (Altelaar et al. 2006; Jackson et al. 2005a)

     

    Fatty acids (Zhang et al. 2007)

     

    Other metabolites

     

     Amino acids (Li et al. 2008)

     

     Flavonoids (Li et al. 2008; Zhang et al. 2007)

     

     Oligosaccharides (Li et al. 2008; Zhang et al. 2007)

     

    Heme (Shimma and Setou 2007; Mazel et al. 2007)

     

    PC phosphatidylcholine, PE phosphatidylethanolamine, PI phosphatidylinositol, PS phosphatidylserine, PG phosphatidyglycerol

     
  2. 2.

    Among the endogenous metabolites, lipids have been intensively investigated. Studies describing detection and imaging of complex lipids (e.g., glycerophospholipids and glycosphingolipids) and simple lipids (e.g., cholesterol, acylglycerides, and fatty acids) have been reported. In addition, other metabolites with superior ionization efficiency such as heme, which is a prosthetic group consisting of an iron atom contained in the center of a porphyrin ring, have also been investigated, even in an African ritual art object (Mazel et al. 2007; summarized in Table 2).

     

We will describe a representative research application in imaging both endogenous and exogenous metabolites.

IMS for exogenous drugs

An important early phase of drug discovery is determining how a candidate drug is distributed and metabolized within the body. The use of IMS to monitor drug delivery has also attracted much interest. Compared to traditional whole-body autoradiography (WBA) using radiolabeled compounds, IMS offers detailed drug distribution images, which are comparable to WBA images. In Fig. 3, Stoeckli et al. show that the two methods produce remarkably similar results. They performed the IMS with WBA using whole-body sections after intratracheal administration of a compound (0.5 mg/kg) to rats. The two corresponding sections were obtained from the same animal but from different regions. Comparison of the methods shows remarkable similarity in the results: high levels are detected in the trachea, the lung, and the stomach while low levels in the blood.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig3_HTML.gif
Fig. 3

Comparison of IMS with WBA using whole-body sections after intratracheal administration of a compound (0.5 mg/kg) to rats. The two corresponding sections are from the same animal but from different positions. Comparison of the methods shows remarkable similarity in the results: high levels are detected in the trachea, the lung, and the stomach and lower levels in blood. Mass spectrometry imaging (MSI) is used as a synonym of IMS (adapted from Stoeckli et al. 2006)

In addition, IMS provides several advantages for determining drug distribution. First, mass-spectrometry-based molecular detection enables simultaneous and discriminative monitoring of both the intact drug molecules and their metabolites (Cornett et al. 2008; Khatib-Shahidi et al. 2006). Figure 4 shows a representative mass spectrum obtained from rats administered the antipsychotic drug olanzapine; both the parent drug and its oxidized metabolite were detected (Fuchser et al. 2008). By its nature, WBA cannot distinguish these molecules. In this regard, IMS can determine whether medicinally intact drugs have reached the target organs or not. Moreover, IMS can visualize the distribution of drugs at a lower cost and in a much shorter time than detection using isotopes (Stoeckli et al. 2006).
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig4_HTML.gif
Fig. 4

Mass spectrum from a single pixel showing simultaneous detection of the mass peak corresponding to (c/t) olanzapine (m/z 313) and its oxidized metabolite (m/z 329) in rat kidney

Figure 5 illustrates the detection of drugs that have been delivered orally in rats (Khatib-Shahidi et al. 2006). Khatib-Shahidi et al. have successfully investigated the distribution of olanzapine and its metabolites in a whole rat sagittal section 2 and 6 h after administering the dose. This study clearly showed the distinct distribution of intact drugs and their metabolites; the intact drug reached the target organ (the brain), whereas its metabolites were localized in the bladder. Furthermore, the time course of the metabolism of parent drugs into demethylated and hydroxymethylated metabolites was visualized over the whole-body section. In this study, notable decreases of olanzapine were observed everywhere in the body except the testis and bladder, while the metabolized compounds accumulated in the bladder. As this example demonstrates very well, IMS-based drug monitoring provides valuable information for drug development.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig5_HTML.gif
Fig. 5

Detection of the drug olanzapine and its metabolite distribution in a whole-body sagittal tissue section by single IMS analysis. Optical images of tissue sections from rats 2 and 6 h after treatment with olanzapine, across four gold MALDI target plates (A). Organs are outlined in red, and pink dot is used as a time point label. MS/MS ion image of olanzapine (m/z 256; B). MS/MS ion image of N-desmethyl metabolite (m/z 256; C). MS/MS ion image of 2-hydroxymethyl metabolite (m/z 272; D). Bar denotes 1 cm (adapted from Khatib-Shahidi et al. 2006)

IMS for endogenous metabolites: phospholipids

Establishment of IMS application for endogenous small metabolites also benefits the fields of toxicology and pharmacology because it will provide an insight into metabolic changes linked to drug administration that are unwanted and deleterious side effects or toxicity. This research field is expanding at an increasing rate, and IMS methodologies for various types of metabolites are developing. Below, we summarize IMS applications for endogenous metabolites.

MALDI-IMS for profiling (Jackson et al. 2005b, 2007a; Jones et al. 2006; Rujoi et al. 2004) and visualizing distribution (Garrett et al. 2006; Hayasaka et al. 2008b; McLean et al. 2007) of endogenous lipids such as phospholipids is the best established application; these lipids can also be imaged with secondary ion mass spectrometry (Colliver et al. 1997; Monroe et al. 2005; Ostrowski et al. 2004; Touboul et al. 2005). This is because phospholipids are ionized efficiently for the following reasons: first, a large fraction—for example, more than 60% by dry weight—of brain tissue consists of lipids. Second, these compounds have an easily ionizable structure; phospholipids, particularly phosphatidylcholines (PCs), contain a phosphate group and trimethylamine that are easily charged (Pulfer and Murphy 2003).

Glycerophospholipids comprise a large molecular family in which phosphoric acid is esterified to a glycerolipid. They are subdivided into distinct classes (e.g., PCs, phosphatidylethanolamines, and phosphatidylinositols) based on the structure of the head group linked to the phosphate, attached at the sn-3 position of the glycerol backbone. They are further subdivided into numerous molecular species on the basis of the composition of the fatty acids linked to the sn-1 and sn-2 positions of the glycerol backbone (Fig. 6). Using IMS, we can image not only these multiple classes but also related molecular species simultaneously. In particular, the capability to determine the distinct localization of each molecular species, that is, to elucidate the distinct fatty acid composition of biological membranes in different tissue locations, is an important advantage of IMS (Fig. 7). Since several types of fatty acids, especially polyunsaturated fatty acids (PUFA), in the phospholipids are released and converted in response to extracellular stimuli into bioactive lipids that mediate important biological processes (Murakami et al. 1997), information on the distinct distributions of PUFA-containing molecular species is quite valuable (Piomelli et al. 2007; Sugiura et al. 2009). Figure 7b shows the distribution of distinct PC molecular species in mouse brain. MALDI-IMS successfully revealed the region-specific distribution of PC molecular species, including PUFA-containing PCs (Fig. 7). Currently, only IMS can visualize such different structures of phospholipids that are hidden in the cellular membrane. Therefore, imaging of individual phospholipid molecular species with IMS will contribute greatly to the field of lipid biochemistry.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig6_HTML.gif
Fig. 6

Structure of phospholipids. Structure of the glycerol backbone of phospholipids (a) is shown. Phospholipids are subdivided into distinct classes (e.g., phosphatidylcholines, phosphatidylethanolamines, and phosphatidylinositols) based on the structure of the head group linked to the phosphate, attached at the sn-3 position of the glycerol backbone (b). They are further subdivided into numerous molecular species on the basis of the composition of fatty acids linked to the sn-1 and sn-2 positions of the glycerol backbone

https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig7_HTML.gif
Fig. 7

a, b Distinct localization of phospholipid molecular species revealed by IMS. Different distribution pattern of phospholipids arise from the distinct fatty acid composition of glycerophospholipids (partially adapted from Sugiura et al. 2009)

IMS for endogenous metabolites: gangliosides

Gangliosides are glycosphingolipids consisting of monosialylated to polysialylated oligosaccharide chains of variable lengths attached to a ceramide unit. They are inserted in the outer layer of the plasma membrane, with the hydrophobic ceramide moiety acting as an anchor while their oligosaccharide moiety is exposed to the external medium (van Echten and Sandhoff 1993). Gangliosides also comprise a large family; their constituent oligosaccharides differ in glycosidic linkage position, sugar configuration, and the contents of neutral sugars and sialic acid. Along with the oligosaccharide unit, the ceramide moiety of gangliosides also varies with respect to the type of long-chain base (LCB; sphingosine base) and the fatty acid to which it is coupled (Fig. 8).
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig8_HTML.gif
Fig. 8

Structures of ganglioside molecular species containing C18-long-chain backbone (LCB) and C20-LCB. The C20 species has two more carbon atoms in its LCB moiety than does the C18 species (arrow)

Previous biochemical studies have revealed that the LCB of the brain ganglioside species has either 18 or 20 carbon atoms (i.e., C18- or C20-sphingosine), and C20-sphingosine (C20-LCB species) is present in significant amounts only in the central nervous system (Jungalwala et al. 1979; Sambasivarao and McCluer 1964; Schwarz et al. 1967; Sonnino and Chigorno 2000). Its content increases significantly in rodents and humans throughout life (Mansson et al. 1978; Palestini et al. 1990, 1991). The C20-LCB gangliosides are of great interest because of their characteristic brain specificity and their dramatic increase during the organism’s life span. However, lack of a visualization technology for specific detection and visualization of C18 and C20 gangliosides has left us incapable of determining their precise tissue distribution. Antibodies to some oligosaccharide moieties are available for visualizing the molecular species with different constituent oligosaccharides (Kotani et al. 1993), but such immunological methods cannot detect the differences in the ceramide structure hidden in the lipid bilayer.

Due to the negative charge on the sialic acids and their rich abundance in the brain, gangliosides are easily detected in the 1,500 < m/z < 2,500 range with IMS in the negative ion detection mode (Chen et al. 2008; Jackson et al. 2005a; Sugiura et al. 2008; Table 3). In addition, IMS discriminates not only structural differences in oligosaccharides but also in the lipid moiety and, therefore, successfully reveals the specific distribution of the C20-LCB species in the mouse brain. While the C18 species is widely distributed throughout the frontal brain, the C20 species is selectively localized in the molecular layer (ML) of the dentate gyrus (DG). Furthermore, the developmental- and aging-related accumulation of the C20 species in the ML-DG can be visualized (Fig. 9; Sugiura et al. 2008), i.e., we can identify the tissue location where C20 gangliosides accumulate. These observations indicate that this brain-region-specific regulation of LCB chain length is, in particular, important for its distinct function in the brain. As this study clearly demonstrates, the novel capabilities of IMS could shed light on long-standing questions in the biological/clinical field.
Table 3

Detection of gangliosides in mouse hippocampus

 

Negative ions

[M-H]

[M + Na-2H]

[M + K-2H]

[M + 2Na-3H]

[M + Na + K-3H]

[M + 2 K-3H]

GM1 (d18:1/18:0)

1,544

GM1 (d20:1/18:0)

1,572

GD1 (d18:1/18:0)

1,858

1,874

GD1 (d20:1/18:0)

1,886

1,902

GT1 (d18:1/18:0)

2,170

2,186

2,202

GT1 (d20:1/18:0)

2,198

2,214

2,230

https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig9_HTML.gif
Fig. 9

Developmental- and aging-related accumulation of C20-GD1 in the ML and SLM of the hippocampal formation. Visualization of the ion corresponding to GD1 (m/z 1,874 and 1,902) in the mouse hippocampus at the indicated time points (postnatal 0, 3, and 14 days and 1 and 33 months). In the P14 mouse hippocampus, C20-GD1 was concentrated in the narrow area of DG-SMm and began to spread over the medial edge of the region (arrow heads). In contrast, the concentration of the C18 species decreased in the ML/SLM with aging (arrows; adapted from Sugiura et al. 2008)

Current limitations and perspective

Despite the promising capability of IMS for imaging small metabolites, there exist problems in this emerging technique at each step of the measurement. These issues could be subdivided into problems concerning (1) sample preparation or (2) MS measurement (summarized in Table 4). We now discuss each of the problems.
Table 4

Current limitation and perspective of IMS for small molecules

Limitation process

Problems

Suggested solutions

Sample preparation (ionization)

High ionization efficiency of the targeted molecule is required

Optimization of sample preparation procedure (Prideaux et al. 2007)

Discovery and development of novel matrices (Astigarraga et al. 2008; McLean et al. 2007) and ionization method (Li et al. 2007, 2008; Taira et al. 2008)

 

MS measurement (ion separation)

Multiple compounds often share the same nominal mass in low-m/z region

Imaging in tandem MS mode (Garrett et al. 2006; Khatib-Shahidi et al. 2006)

Ion separation with ion mobility cell (Jackson et al. 2007a; McLean et al. 2007)

Imaging with FTICR instrument (Cornett et al. 2008)

Development of organic matrix-free ionization method (such as use of nanoparticles (Jackson et al. 2007a; Taira et al. 2008), ME-SALDI (Liu et al. 2008), DIOS (Liu et al. 2007), and NIMS (Northen et al. 2007))

  1. 1.

    The sample preparation procedure must be optimized to maximize detection efficiency of the targeted molecule.

     

It is necessary to optimize several parameters of the sample preparation procedure including matrix selection and solvents used to extract analytes from tissues, according to the physical and chemical properties of the targeted molecule. In fact, previous studies have shown that, due to the nature of MALDI, successful detection of different molecules of interest can depend on the sample preparation procedure for the following reason. Since tissues and cells are the subjects of MALDI-IMS, the sample cleanup procedure is limited, whereas, in traditional MS, analyte molecules are generally extracted and separated from crude samples by gas chromatography or high-performance liquid chromatography. When such a crude sample is subjected to MS, numerous molecular species compete for ionization, and molecules that are easily ionized reach the detector preferentially while suppressing the ionization of other molecules, causing severe ion suppression effects (Annesley 2003; Gharahdaghi et al. 1996; Krause et al. 1999). In fact, using a mouse whole-body section coated with analyte drugs, Stoeckli and colleagues have found small regions (although less than 5% of the total region) in which the drugs could not be detected, presumably because of ion suppression effects (Stoeckli et al. 2006). Thus, optimizing the sample condition so that the analyte molecule present in the crude mixture can be efficiently ionized is an important issue. For this purpose, sample preparation has a critical role. In particular, it is helpful to perform a preliminary experiment using a reference drug because the choice of a suitable matrix compound, as well as optimizing the composition of the matrix solution, can improve the ionization efficiency of the molecules of interest.

Atkinson and colleagues have described an excellent example of this problem. They tried to visualize the distribution of an anticancer compound, AQ4N, but found that it is hardly ionized by itself because it forms complexes with DNA. By screening matrices, they found one that could support the ionization of AQ4N without generating complexes by raising the acidity of the matrix solution and succeeded in visualizing the distribution of AQ4N (Fig. 10; Atkinson et al. 2007). Thus, important tasks for the future are discovery and development of novel matrices (Astigarraga et al. 2008; McLean et al. 2007) and alternative ionization methods (Li et al. 2007, 2008; Taira et al. 2008).
  1. 2.

    Advanced MS analysis methods should be applied to separate multiple compounds with the same nominal mass.

     
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig10_HTML.gif
Fig. 10

Optimization of the ionization conditions for the drug banoxantrone (AQ4). Since AQ4 forms complexes with DNA, a high concentration of trifluoroacetic acid (TFA) was added to the matrix solution to avoid complex formation (adapted from Atkinson et al. 2007)

The low-m/z region (m/z <1,000) of a MALDI spectrum contains a large population of ions from endogenous metabolites, as well as matrix-related adduct clusters and fragments (Cornett et al. 2008; Garrett et al. 2006). Because of such a high density of ions, multiple compounds often share the same nominal mass. In such cases, an effective solution is to perform advanced MS analysis to distinguish the target compounds of interest from this extensive chemical background. Several research groups have demonstrated that imaging in the tandem-MS mode provides highly specific information about a target molecule. Essentially, all ions in the range of the targeted precursor are fragmented, and the abundance of the target compound is determined from the measured intensity of one or more of its structurally significant fragment ions. This measurement is performed in the tissue for each precursor of interest (Garrett et al. 2006; Khatib-Shahidi et al. 2006). By performing tandem-MS scanning, Kowalski et al. successfully generated highly selective drug distribution images (Fig. 11). Furthermore, the combination of tandem MS and ion mobility separation adds a further separation step and provides more specific information (Jackson et al. 2007a; McLean et al. 2007; Trim et al. 2008).
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig11_HTML.gif
Fig. 11

Specific detection of olanzapine drug by imaging in the tandem MS mode. a MALDI-TOF/TOF spectra of mass peak corresponding to olanzapine drug (m/z 313). A major fragment mass peak was observed at m/z 256. ba, c Optical image of a histological stained 20-µm-thick section. b The m/z 256 fragment ion image has been generated indicating the specific location within the tissue of the parent olanzapine drug by selective reaction monitoring. d The overlay image combines the optical image with the molecular ion image to illustrate the localization of the drug (adapted from Kowalski et al. 2007)

Another approach is to use MALDI–Fourier transform ion cyclotron resonance (FTICR) MS for tissue imaging (Cornett et al. 2008). Cornett et al. showed that this technique utilizes its high resolving power to produce images from thousands of ions measured during a single MS pass. Accurate mass measurement provides high molecular specificity for the ion images on the basis of elemental composition. Figure 12 shows FTICR images of hydroxymethyl olanzapine and three isobaric ions (Cornett et al. 2008). This example clearly demonstrates that the four major ions were observed at the same nominal m/z as the 2-hydroxymethyl metabolite of olanzapine (exact mass 329.1431). Although they were detected within a range of only 0.02 u, these ions were sufficiently resolved by FTICR MS, and each ion exhibited a distinct ion distribution image.
https://static-content.springer.com/image/art%3A10.1007%2Fs11481-009-9162-6/MediaObjects/11481_2009_9162_Fig12_HTML.gif
Fig. 12

MALDI-FTICR images of hydroxymethyl olanzapine and three isobaric ions. The four major ions observed at the same nominal m/z as the 2-hydroxymethyl metabolite of olanzapine (exact mass 329.1431). Although they were detected within a range of only 0.02 u, these ions are sufficiently resolved by use of FTICR MS and each ion exhibited a distinct ion distribution image (adapted from Cornett et al. 2008)

We also note that elimination of matrix-derived ions is also effective for reducing the overlap of mass peaks from multiple compounds, thereby obtaining specific information in a targeted analysis. To achieve this, several important studies have been published in which organic matrix-free ionization was developed, such as the use of nanoparticle-based ionization (Jackson et al. 2007a; Taira et al. 2008), matrix-enhanced surface-assisted desorption/ionization (Liu et al. 2008), desorption/ionization on silicon (Liu et al. 2007), and nanostructure-initiator mass spectrometry (Northen et al. 2007).

Conclusion

IMS on small molecules has opened a new frontier in pharmacology and toxicology. As discussed above, MALDI-IMS provides an attractive alternative for monitoring drug distribution within animal models, with much faster results and lower cost than the traditional WBA method. Furthermore, its unique capability to simultaneously image many types of molecules enables distinct visualization of a parent drug and its metabolites. This advantage is important for pharmacological and toxicological study because we can learn whether the intact compound reaches desirable or undesirable organs.

As another application, we can also localize numerous endogenous metabolites such as lipids; this will benefit not only biological research but also pharmacological research since metabolite reactions in response to drug administration may be assessed.

Herein, we have discussed the promising capabilities of IMS, as well as the importance of the experimental protocol, particularly the sample preparation step. As introduced here, with attention paid to some technical points, MALDI-IMS provides valuable information for exploring the in situ pharmacometabolome that could not be obtained by any other existing technique: high selectivity, rapid acquisition, and parallel acquisition of multiple analytes. We expect that continued improvement in the experimental protocol, as well as in the MS instrumentation, will further expand the capability of this emerging technique. In conclusion, we hope this review will help readers to explore IMS as a new tool in their research fields.

Copyright information

© Springer Science+Business Media, LLC 2009